Frogs O' War - 2019 Stats O’ War Football Preview SeriesThe #1 TCU Athletics blog on the internet!https://cdn.vox-cdn.com/community_logos/50293/fow-fav.png2019-08-22T11:00:00-05:00http://www.frogsowar.com/rss/stream/184572222019-08-22T11:00:00-05:002019-08-22T11:00:00-05:00Stats O’ War: Does TCU Rush Too Much?
<figure>
<img alt="" src="https://cdn.vox-cdn.com/thumbor/_g1Rc-5SRllWBywgHNgCfQmyXHw=/0x264:1584x1320/1310x873/cdn.vox-cdn.com/uploads/chorus_image/image/65090169/SOW.0.jpg" />
</figure>
<p>As the season starts, we can use expected points to understand how TCU’s offensive changes might lead the Frogs to Big 12 contention.</p> <p id="wEygXr">Earlier this offseason at Football Study Hall, <a href="https://www.footballstudyhall.com/2019/5/31/18647463/tcus-next-offensive-adjustment-veer-and-shoot-matthew-baldwin-gary-patterson">Ian Boyd</a> detailed some subtle changes coming to the TCU offense. With the addition of WR coach Malcom Kelly and the emergence of Taye Barber as a true #2 threat in spring ball, TCU looks to move their passing game downfield, incorporating some of the more dynamic and vertical offensive concepts the Big 12 has become accustomed to. </p>
<p id="A8dRjV">In fact, Gary Patterson himself has hinted at this new vertical offense: the coach has admitted that TCU is throwing the ball downfield as well as they ever had and touted the quality of competition at the quarterback spot. Whoever wins the starting job (my bet: Delton takes the first snap), it appears that TCU’s offense will be geared more towards deep threats than it’s been in the past. </p>
<p id="H7Kf9t">But wait, TCU returns five starters on the offensive line and a three-headed monster backfield of upperclassmen; surely, the team whose rushing attack outperformed their passing attack, returning so much, would want to increase rushing volume? </p>
<p id="9Psf6a">Let’s take this idea head on, using 2018 play-by-play data and the nifty EPA data <a href="https://www.frogsowar.com/2019/8/13/20802128/expected-points-big-12-college-football-analytics">I introduced last week</a>: <em><strong>Does TCU rush too much?</strong></em></p>
<p id="OftU35">Each game, a football team implicitly faces a simple optimization problem. A team calls plays to maximize expected value subject to a couple simple constraints. For our purposes, the team is choosing a mix of run and pass plays to maximize expected value (EPA per play * share of plays), subject to the fact that the share of run and pass plays must equal one (that is, you must pick a either a run or pass every play). Does this oversimplify? Yes, but the simplification meaningfully isolates an important facet of decision-making. </p>
<p id="3IC9tX">Football and analytics nerds on Twitter will tell you that early down rush rates are indicative of a lacking passing game, and that teams too often rely on the rush. Let’s see how that bears out in college football: </p>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/SdaPP6PUkhUZ7QGjtMIUKP_3E9w=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/19056701/rush.pass.rates.png">
</figure>
<p id="7oe0Fz">The linear fit, the blue line, is downward sloping, indicating a negative relationship between early down rush rate and passing success rate: teams who rush on early downs more often are worse at passing. That’s probably more descriptive than causal, but still provides some guidance into proceeding towards answering our question: early down rush rate matters for the passing game. </p>
<p id="RxGV3k">TCU lies directly along the trend line above, just over 45% in rushing success and just under 45% in passing success, middle of the pack in both categories. On all downs, TCU’s offense ranked 86th in rushing efficiency, 113th in marginal explosiveness for rushing (101st, 71st for passing). TCU was more successful on the run than the pass last year, in terms of efficiency, but in terms of expected points added, <em>the opposite is true. </em></p>
<h3 id="yzq1I2"><strong>Methodology</strong></h3>
<p id="SM1Rwq">I’ve filtered 2018 college football play-by-play data from <a href="http://CollegeFootballData.com">CollegeFootballData.com</a>, limiting observations to first and second downs, between the 20s, not in garbage time. That’s a severe restriction on an already limited sample, for sure, but it provides the clearest picture of team style by stripping away context that dictates what teams do. Of course, we don’t want TCU to pass more on third and short, or run more on third and long. Early downs and close games, out of the red zone/backed up zone generate the clearest picture of a team’s decision-making.</p>
<p id="lZuR11">I constructed an “expected value” for a run and a pass for TCU. Simply put, it’s just the average EPA from a run times the rush rate plus an average EPA from a pass times the pass rate: </p>
<blockquote><p id="dqO7MO">Expected Value = (EPA_rush * Rush %) + (EPA pass * Pass %)</p></blockquote>
<p id="wZR23B">TCU ran 453 plays in between the 20s, against FBS opponents, in non-garbage time this season. On average, the expected value of a play in this sample for TCU is (-.0146 * .42) + (.156 * .443) = -.006132 + .069108 = <strong>.063 expected points added</strong>, slightly positive. </p>
<p id="wWp8Su">Multiple approaches lend an answer to the question of whether TCU rushes too much. First, given the relative EPA of a rush and a pass (-0.146 for a rush, +0.156 for a pass) for TCU’s 2018 offense, the naive answer is: yes, TCU should pass more, because the expected value of a pass is higher than a rush. This is a naive approach, <strong>but it’s not a bad approach! </strong>There’s definitely an argument here. </p>
<p id="yvnanu">Let’s go a little deeper, and incorporate success rates, because, after all, one might believe passing is a riskier enterprise. </p>
<p id="xZVoN9">Consider a dramatic increase in passes, say 15%, which corresponds to about 68 additional passes over the course of the season (17 a game) for TCU. Using this naive approach, TCU would replace 19.49 expected points added with 42.17 EPA, a net of +22.663 across the season. That’s a non-trivial swing! </p>
<p id="akMqZj">We can make this analysis more robust by factoring in the mean loss of EPA on a rush and a pass, which will round out what might happen when we change strategy a little more fully. The formula we’ll use is:</p>
<blockquote><p id="NkRkAH">EXPECTED VALUE = 68 * [EPA(success pass) * (sr_pass) + EPA(fail pass) * (1-sr_pass)] - [EPA(success rush) * (sr_rush) + EPA(fail rush) * (1-sr_rush)]</p></blockquote>
<p id="ztPLHN">In words, we are subtracting the difference between the Expected Value of EPA on a pass and on a rush, and multiplying that out by 68 plays over the course of the season. Using this method, by rushing 15% more, TCU would’ve replaced -12.92 expected points added in the run game with +4.44 EPA in passes, a swing of almost 17 points!</p>
<h3 id="PbxKc4"><strong>Aw Heck, Man</strong></h3>
<p id="ne9UeP">The shrewd reader will be asking, about this point, “this is all good and well, but we’re taking success rate as a given; wouldn’t a change in run/pass tendencies affect the success rates of runs and passes?” </p>
<p id="wqlu45">Here, I’ll employ perhaps the simplest application of the <a href="https://en.wikipedia.org/wiki/Heckman_correction">Heckman Selection model</a> you’ll see in the wild. The Heckman model just addresses the idea of non-random selection; here, I’ll assume a “worst case scenario” replacement of runs with passes, assuming most of the rushes we replaced were good rushes, and the passes we replaced them with are bad passes.</p>
<p id="0HUYQq">Being generous, let’s say a 15% decrease in rushes, all of them “bad rushes”, would increase success rate on rushes to 45% (which is top 30 in the nation, so yes, this is an absurd generosity), and more passing decreases success rate to 40% (which would move TCU into the bottom third of teams nationally, again, absurd). Under this ridiculous circumstance, TCU would replace .3042 + -.45485 = -.15 * 68 = -10.244 expected EPA for rushing with .56 - .597 = -.037* 68 = -2.516 expected EPA for passing, which is still almost an 8 point swing!</p>
<p id="9GqnkX">Even in the worst case scenario for passing and the best case scenario for rushing, an increase in passing rates represents an absolute improvement in expected points outcomes!</p>
<h3 id="g1QCsZ"><strong>Taking it to the Games</strong></h3>
<p id="mIRGK5">So, of course, a (conservatively-estimated) increase 17 expected points added doesn’t translate to 17 actual points spread out across the season, but it should translate to some positive increase in points across numerous situations. Looking at individual games can highlight specific points in the season where TCU was rushing too much.</p>
<p id="eMYUrU">First, let’s start with the two one-score losses: vs. Texas Tech and at Kansas. </p>
<p id="K0aAdu">In the Tech game, the expected value of a rush for TCU was -.340 EPA, and the Frogs rushed on 35.4% of plays in the sample, passing 64.6% of the time. You might think, <em>“wow, that’s a lot of passes!” </em>and you’d be right: 65.6% was TCU’s highest pass rate in a game in the sample. But 64.6% passing rate <em>still wasn’t enough</em>. The expected value of a pass for TCU against Tech was +0.0271, and by passing just 10 more times in the game, TCU could’ve improved it’s expected EPA by a net +3.67 points! The increase in expected points doesn’t mean TCU would’ve won the game or pulled 3.67 points out of the air to beat Tech, but rather means that in a one score game, in which the Frogs had the ball on the final possession, TCU left 3.67 points worth of expected value on the table by rushing too much on early downs. </p>
<p id="pBPRL6">Against Kansas, the expected value of a rush was -0.0473 points, and the expected value of a pass was +1.11 points (shout out Michael Collins and Kansas’s bad pass defense). Given rush and pass rates of 62.5% and 37.5%, respectively, increasing early down passes by just 10 would’ve increased expected EPA by 11.6 points! Again, that doesn’t directly translate to points, but it does translate to increased opportunities to score points against a terrible defense, and more opportunities to avoid the most embarrassing loss in college football.</p>
<p id="yzcbl0">Here are the other games where TCU “over-ran”, and could’ve strictly improved their expected outcomes by simply increasing their pass volume:</p>
<ul>
<li id="c72M76">Iowa State (+2.46 expected EPA from ten additional passes)</li>
<li id="mev7RX">Kansas State (+5.45)</li>
<li id="j9AWBa">Ohio State (+7.66)</li>
<li id="oTjp6S">Oklahoma (+4.66)</li>
<li id="Pv9DAW">West Virginia (+1.02)</li>
</ul>
<h3 id="skYiaQ"><strong>Conclusion</strong></h3>
<p id="nX4gBj">Multiple methods of inquiry confirm the simple fact:<strong> TCU consistently left scoring opportunity on the field by rushing too much on early downs last season</strong>. An increase of 10 passes in TCU’s two one-score losses could swung their expected EPA by more than 3 points (vs. Tech) and more than 11 points (vs. Kansas). </p>
<p id="rdvW1Z">As we enter the 2019 season, TCU’s offense remains a bit of a question mark. For TCU to return to its status as a contender in the Big 12, they’ll have to optimize talent and shake off their offensive woes. As the new downfield threat is rolled out, we will keep tabs on how rush rates change from last year, and with that, how TCU’s expected EPA benefits from offensive adjustments. </p>
https://www.frogsowar.com/2019/8/22/20814150/stats-o-war-does-tcu-rush-too-muchstatsowar2019-08-13T13:00:00-05:002019-08-13T13:00:00-05:00Expected Points: A Better Way to Evaluate College Football Performance
<figure>
<img alt="" src="https://cdn.vox-cdn.com/thumbor/UIm1vqgUvMCvxZB5k2xJexT5gPE=/0x264:1584x1320/1310x873/cdn.vox-cdn.com/uploads/chorus_image/image/65010246/SOW.0.jpg" />
</figure>
<p>Parker examines a new way to dive into advanced stats.</p> <p id="SKcFGm">College football data is entering a new age, my friends. Thanks to some hard work from <a href="http://collegefootballdata.com/">collegefootballdata.com</a> and my new online friend <a href="https://twitter.com/903124S">@903124</a>, we have a reliable, useful, and pretty darn consistent measure of Expected Points Added (technical details <a href="https://github.com/903124/CFB_EPA_data">here</a>).</p>
<p id="aErsWv">This is great news! Now, we can start to talk about plays in context, and fine tune a metrical evaluation and comparison of teams. First, I’ll explain EPA and highlight how it is informative. Then, I’ll look at some TCU performances in 2018 according to EPA, and in doing so, I’ll lay out an analytical framework for the upcoming season using EPA.</p>
<h2 id="67dkvn"><strong>Expected Points Added, by the Numbers</strong></h2>
<p id="h38SkE">An expected points metric allows us to quantitatively evaluate a play, attaching a specific magnitude and direction to each play’s effect on a game. The EPA metric is derived using a model that breaks plays into down, distance, and yard line bins, and then estimates the actual points expected from each state, using some fancy math. </p>
<p id="LvK5A9">Expected points actually explains pretty well, at face value, what happened on a play. For example:</p>
<p id="jcK3tO"><strong>Top five plays by EPA, 2018: </strong></p>
<ol>
<li id="lflDlo">UAB vs Rice, 2nd Quarter. 3rd and 6 from their own five, up 21 points: UAB throws a 95 yard touchdown pass. Expected Points: -2.19. Actual points: 7. EPA: 9.79. </li>
<li id="KlH7zO">Notre Dame vs Va Tech, 3rd Quarter, 2nd and 12 from their own 3, up 8 points: Notre Dame scores a 97 yard rushing touchdown. Expected Points: -2.64. Actual Points: 7. EPA: 9.26.</li>
<li id="lZqVoQ">Colorado vs Colorado St, 3rd Quarter, 3rd and 14 from their own 11, up 25: Colorado scores an 89 yard touchdown pass. Expected points: -2.18. Actual Points: 7. EPA: 9.18. </li>
<li id="oufUp3">Minnesota vs Illinois, 4th Quarter, 3rd and 19 from their own 14, down 24. Minnesota throws an 86 yard touchdown pass. Expected Points: -1.95. Actual Points: 7. EPA: 8.95.</li>
<li id="IJqsxJ">UTEP vs. NMSU, 3rd Quarter, 3rd and 18 from their own 24, tie game: UTEP Throws a 76 yard touchdown pass. Expected Points: -1.79. Actual Points: 7. EPA: 8.75. </li>
</ol>
<p id="QuY8tu"><strong>Top five TCU plays by EPA, 2018</strong></p>
<ol>
<li id="jONbjQ">vs Ohio State, 2nd quarter, 2nd and 10 from the 7 yard line, down 4: Darius Anderson runs for a 93 yard touchdown. EP: -0.8 AP: 7 EPA: 7.8</li>
<li id="guWLe3">vs Baylor, 2nd quarter, 3rd and 6 from the 35: the Mule hits Reagor, who runs for 65 yard touchdown. EP: -0.6 AP: 7. EPA: 7.06.</li>
<li id="2smz3n">vs Oklahoma State, 3rd quarter, 1st and 10 from the 17, up 18: Jalen Reagor runs 83 yards for a touchdown. EP: .107. AP: 7. EPA: 6.89</li>
<li id="aLihkh">vs Kansas State, 3rd quarter, 2nd and 10 from the 33: Collins hits Reagor for a 67 yard touchdown. EP: .392 AP: 7. EPA: 6.6.</li>
<li id="SWvgUH">vs Ohio State, 3rd quarter, 2nd and 11 from the 49: Robinson hits Hights for a 51 yard touchdown. EP: .888. AP: 7. EPA: 6.111.</li>
</ol>
<p id="qUxgfJ">Most of these were situations where a team wasn’t expected to score very often, and they did. Hence a high EPA. This can be fun to play around with, especially when trying to consider context - the EPA doesn’t account for when the margin is close, or when the game is late, but it does allow to compare by summing up across those situations. For example: </p>
<p id="HrdBID">TCU EPA total 2018 (-80.406). <br>TCU EPA second half of games 2018: (-56.65)<br>TCU EPA first half of games 2018: (-23.754)</p>
<p id="ccht2r">Wow - TCU was really bad! They were especially bad in the second half! There’s a wealth of breakdowns we could do - and that I will do - but for now, this is enough to introduce the general idea and kind of check the validity. </p>
<h2 id="kUJWsL"><strong>Using EPA to Evaluate College Football Teams</strong></h2>
<p id="jfvaq4">The NFL Analytics Dark Web (<a href="https://go.redirectingat.com?id=66960X1516590&xs=1&url=https%3A%2F%2Ftheathletic.com%2F896789%2F2019%2F05%2F16%2Finside-the-nfl-analytics-dark-web%2F&referrer=sbnation.com&sref=https%3A%2F%2Fwww.frogsowar.com%2F2019%2F8%2F13%2F20802128%2Fexpected-points-big-12-college-football-analytics" rel="sponsored nofollow noopener" target="_blank">yes, it’s a thing</a>), loves EPA! It’s a measure of success, of explosiveness, and of player execution, all in one convenient metric. Of course, that adds a bit of murkiness in interpretation, but on the whole, a “mean EPA” stat really tells you a lot about a team, and helps you to compare a team. </p>
<p id="onaeY5">First, a graph: </p>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/1w69rdqH64SEsnm7ScLT23hGnig=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/18962859/EPts.jpg">
</figure>
<p id="QMIqYp">The intuition behind expected points is clear, but the consistent estimation is a bit tricky. This graph demonstrates a pretty smooth expected points trend, with some movement around a straight line. Also, there’s an exponential shape here - expected points increase by a lot (more than linear) the closer you get to your opponents’ yard line. This provides an immediate solution to one problem in college football analysis: <strong>all yards are not created equal. </strong></p>
<p id="EsJQ7b">Using EPA allows us to compare performance across context, punishing teams more for making mistakes in high leverage areas and rewarding them more for excelling in those high leverage areas. </p>
<p id="mkojua">Let’s look at TCU’s season with EPA. First, the offense:</p>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/MWkKvM4bU6jm5QkPilJnyG8-hZ4=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/18962886/TCU_EPA_Offense.png">
</figure>
<p id="CGqdx1">What we are looking at here is Expected Points Added by Yard Line for all of TCU’s offensive plays. Remember, yard line 0 corresponds to your opponents’ end zone, and note that the red line is just a demarcation of 0 EPA. </p>
<p id="BBynd4">You’ll notice immediately a couple of things. First, you’ll see that there are a whole dang lot (technical term) of negative EPA plays, particularly in scoring opportunities. In fact, if we define a scoring opportunity as “inside your opponent’s 40”, a loose definition, we see <strong>ELEVEN</strong> plays with -5 EPA or more. TCU ranked 124th last year in points per scoring opportunity, and in the bottom third of FBS in success rate inside the 30. This graph bears that out. Let’s look at some of those plays in detail to see what happened and how EPA corresponds to on-field activity. </p>
<h3 id="PUhhlS"><strong>TCU Worst Plays, 2018:</strong></h3>
<ol>
<li id="SchxEx">
<strong>Butt-Fumble: </strong>vs. Kansas, 4th quarter, first and goal from the nine. Darius Anderson fumbles as TCU sets up game-clinching field goal. EPA: -11.4.</li>
<li id="Dx3yvy">
<strong>Not Kicking a Field Goal: </strong>vs. Kansas, 1st quarter, fourth and goal from the 1. Instead of kicking a field goal like rational people, the Frogs don’t convert on the fourth and goal try. EPA: -11.05.</li>
<li id="uGOJCW">
<strong>We Passed it to the TE: </strong>vs. Iowa State, 3rd quarter, second and ten from the 14. Pass completed to Artayvious Lynn, who fumbles. EPA: -9.687</li>
<li id="Dc9elX">
<strong>Shawn Robinson (x3 tie): </strong> Interception vs Texas Tech, Interception vs Texas, Interception vs SMU. EPA: -9.36, -8.8, -7.75. These were all bad, no need to pile on. </li>
</ol>
<p id="uP8Jkg">These are pretty much the plays you’d pick off the top of your head! EPA just quantifies that! Four of these plays were effectively giving away a win (Kansas x2, Tech, Texas), and the other two made games a lot more competitive than they perhaps should’ve been. The key takeaway:<em><strong> </strong></em>TCU was <strong>abysmal</strong> in scoring opportunities last year. Their EPA Per Play for downs 1-3 inside the opponent forty yard line was -.349! </p>
<p id="yJoSmK">You heard that right! On average, the value of a TCU play in opponents’ territory was <strong>negative. </strong>Even if you filter out outliers (plays with an EPA < -5), the mean is -.096. That’s... not good. </p>
<p id="5QFbbb">Breaking it down by rush and pass isn’t much better: TCU averaged -.114 EPA on rushes last season (8th in the Big 12) and -.082 on passes (also 8th in the Big 12). </p>
<p id="boijry">The Frogs were the best defensive team in the Big 12 by a full tenth of a point, allowing barely above zero EPA for the season (.009).</p>
<p id="kPYZ4D"></p>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/0BHyMR6oXHLJEpCpAupm1jF-sdk=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/18963051/tcu_defense.png">
</figure>
<p id="tZmEwF">Since it’s defense, remember that TCU’s goal line is the 0, and that negative EPA is good for defense. Here, you see a couple big plays (10 by my count), but also some pretty strong swings - TCU defense had 7 plays of at least -2.5 plays, and as they got closer to their goal line, they really kept teams from exploiting scoring opps. </p>
<p id="ERhSnn">Lastly, let’s examine the offense one more time, in the context of the down by down preview series I did earlier this summer. I’ll filter out possessions with more than a two score margin, and let’s see how TCU did on rushing and passing by down in a neutral game script. </p>
<h3 id="Hxyn9l"><strong>TCU Offense, Neutral Game Script, 2018</strong></h3>
<p id="BeUMVj">Down Total EPA (Rush/Pass)</p>
<ol><li id="uDykVs">-.1079 (-.104/ -.113)</li></ol>
<p id="doC6Cq">2. -.0256 (-.165/ +.129)</p>
<p id="QAtLaH">3. -.2442 (-.057/ -.335)</p>
<p id="0TAjpf">Third downs killed TCU! That’s especially painful considering TCU faced the 8th shortest third down situation last year, on average. The only plus situation for TCU was passing on second down. (For reference, OU’s second down passing attack was +.412). Passing on third downs was basically a non-starter, a little bit of an effect of the short third down distance, but that highlights a real deficiency in TCU’s offense. </p>
<h1 id="6o2GwE"><strong>Conclusion</strong></h1>
<p id="JQG38v">I could keep going with different EPA breakdowns (and I will!) but the above is a good introduction to the concept. As we go through the season, we can quantify exactly how good and bad a play was, and start to attach EPA numbers to specific players and offenses. I’m excited to use EPA throughout the season as a better way to evaluate teams in different contexts. </p>
<p id="xcfpgt"></p>
https://www.frogsowar.com/2019/8/13/20802128/expected-points-big-12-college-football-analyticsstatsowar2019-07-19T12:30:00-05:002019-07-19T12:30:00-05:00Stats O’ War Season Preview: Third Down Offenses
<figure>
<img alt="" src="https://cdn.vox-cdn.com/thumbor/3noQFpGpZl_G0ITcwRA2s1XcMds=/0x60:360x300/1310x873/cdn.vox-cdn.com/uploads/chorus_image/image/64764897/Image_from_iOS__1_.0.jpg" />
</figure>
<p>Third downs are what make or break a drive for most offenses - how do Big 12 teams approach third downs, and how do they succeed? </p> <p id="xF5Y7H"><em>Author’s note: This is the third in a three-part offensive scouting exercise in preparation for the 2019 season, and the third entry in Stats O’ War’s inaugural College Football Season Preview. You can find the </em><a href="https://www.frogsowar.com/2019/6/27/18693180/big-12-preview-offenses-first-down-advanced-stats"><em>First Down Scouting Preview here</em></a><em>, and the </em><a href="https://www.frogsowar.com/2019/7/11/20689566/stats-o-war-season-preview-second-down-offenses"><em>Second Down Scouting Preview here.</em></a></p>
<p id="sKNbxS"><em>A note about data: All data in this post is filtered for garbage time. Additionally, due to coaching turnover, I have proxied each program with the most recent team associated with their offensive manager. When you read “Texas Tech, Kansas, and West Virginia” below, the data are coming from Utah State, Florida, and Troy, respectively. </em></p>
<p id="wMn416"><em>The first part of this post focuses on TCU’s Big 12 opponents. If you’re just here for info on the Frogs, you can skip down to the “What About TCU?” section at the bottom.</em></p>
<hr class="p-entry-hr" id="ci4Yrt">
<p id="E0yTSE">This summer, I’ve been working through scouting and previewing the Big 12 Offenses, trying to glean any information about what we might expect to see on the field this fall, all with the purpose of enhancing our collective football viewing. How we view teams really helps us to contextualize statistics and understand more of the workings of the game. So far, I’ve looked at first downs, isolating the first play of a possession to analyze style, and at second downs, tracing out which teams are making up ground on second down and which teams are taking shots to get ahead. </p>
<p id="hZqs2C">Third downs are the result of a dynamical process of first and second down results. A team’s success might mean they don’t see a third down on a drive, or a team’s failure might mean they see a third down every drive. Some teams are committed to low ceiling, and see themselves in third-and-manageable often. Other teams are boom and bust, finding themselves in short and long third down situations equally. </p>
<p id="Gocgcn">Third down can be a key indicator of a team’s quality, but it serves as an even better bellwether of what teams do when they don’t succeed on first and second downs. In this installment of the Stats O` War Season Preview, I analyze third downs in 2018 to characterize what situations teams find themselves in and how they approach those situations.</p>
<h1 id="dP3Mic"><strong>Third Downs: The Basics</strong></h1>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/CldRyLGZeb_77HebDuoW9jpmbc8=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/18321067/thirddownsuccess.jpg">
</figure>
<p id="ABqEFc">I began the analysis with a basic visualization of the situation teams found themselves in most often, and their success rate overall. The top left quadrant reflects an average distance faced on third down lower than average and a higher than average third down success rate. The bottom right reflects a higher than average third down distance and a lower than average success rate. A few key features of this graph jump out, as we initially characterize the third down landscape of Big 12 teams. </p>
<p id="v4ngYo">First, we note the effective “peak” of Texas Tech, Texas, and OU. Those three teams sat close to the mean of average third down distance, yet all sit comfortably above average in third down success. Texas, as we’ve seen in the last few posts, is a little different in their offensive approach - they rushed and self-imposed a low ceiling on first and second down, committing to a three-down approach. That’s reflected here, where their average third down is lower than half the conference, and the Longhorns were efficient at getting the final piece to continuing drives - 46.1% of the time, Texas converted on third downs. </p>
<p id="KnrSzO">As for Texas Tech and OU, we have two offensive specialists (Matt Wells and Lincoln Riley) lead by two experienced and elite QB talents (Jordan Love and Kyler Murray). The experience shows, as the two offenses were able to convert on third downs absent any kind of positional advantage. This fall, one could expect Matt Wells’s play-calling to come through and continue some of that third down success, and QB Alan Bowman has showed promise in the past - the combination of an intelligent offense with some stability at QB could give Texas Tech a secret weapon in third down conversions. </p>
<p id="IkGmfe">Oklahoma, of course, continues their long line of quarterback succession with proven starter Jalen Hurts, whose running ability should continue to make OU’s offense yet again a unit that allows their opponents no rest. </p>
<p id="iqA3ww">Iowa State and West Virginia both got themselves in trouble on third downs - both found themselves, on average, facing a third down and over seven yards, and as a result, neither was particularly effective. The other big split comes from your pass-heavy and your rush-heavy teams: Baylor and Oklahoma State, taking more passing shots on first and second down, find themselves in longer situations than rush-heavy Kansas and Kansas State. While the differences in distance among those two groups reflect the stylistic variations of those offenses, their relative effectiveness was about the same - all four teams made up the Big 12’s Middle Class of third down efficiency. </p>
<p id="utUtfJ">To continue an evaluation of team performance on third down, I now turn to rush rates. We know which teams found themselves in longer third down situations (Oklahoma State, Baylor, WVU, Iowa State) and which were most successful (Oklahoma, Texas Tech, and Texas), and we can examine who relied on the rush versus the pass most often to find their success. </p>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/mt2ZD24tTMZLnWIyA4IjsnuZPcw=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/18321092/thirddown_rush_rate.jpg">
</figure>
<p id="O95rol">Four of the top five teams in success rate passed more than league average. Kansas and Texas sit as the lone members of the “above average rush rate, above average success” quadrant. Texas, and to some extent Kansas (proxied by Dan Mullen’s Florida) both love the QB run, whether by design or by opportunity. Texas’s above average rush rate is probably the Ehlinger factor - Kansas won’t off the bat have a QB rush option anywhere near that caliber, but Pooka Williams has proved a dangerous rusher, and third downs might be where he adds the most marginal value to an unknown Kansas offense. Neal Brown’s entrance into the Big 12 is already making waves - seeing West Virginia near the top in rush rate is a little shocking, but that balance is not the least of the reasons WVU brought him in. </p>
<h1 id="TWLybr"><strong>Context and Situational Third Downs</strong></h1>
<p id="7aOgWf">Of course, not all third downs are created equal. Let’s break up the data by situation - third and short (<4), third and medium (<7), and third and long (>7) - and see who excelled where. </p>
<h3 id="01PgVD"><strong>Third and Short: </strong></h3>
<p id="XBLkWe">Top 5 Success Rate, 3rd and Short:</p>
<ol>
<li id="TQstSA">West Virginia (84.2%)</li>
<li id="EJauLp">Iowa State (81.2%)</li>
<li id="i0QHBK">Texas Tech (77.8%)</li>
<li id="Ac0YzG">Baylor (77.5%)</li>
<li id="u1ry9q">Texas (68.8%)</li>
</ol>
<p id="WhgtEf">It shouldn’t surprise us to see West Virginia in this list - these teams are all similarly modern, emphasizing a running quarterback and far more balance than the prototypical Big 12 spread. Neal Brown and Matt Wells, particularly, are much more in the mold of the “evolved spread” and bring vicious running attacks with them, pointing to a new fact of reality in the Big 12: you can’t win without a quarterback to make plays and a running attack to keep your opponents honest. A small aside: one could argue that some second tier contenders who haven’t made the jump thus far only had one facet - 2018 Oklahoma State and TCU only had the rushing attacks, it took Iowa State a couple weeks to pair their QB with a running attack - and the 2018 Big 12 surprise, Texas, had a quarterback and a running back pair the likes of which they haven’t seen in years. </p>
<p id="08JSIg">Top 5 Success Rate, Third and Medium:</p>
<ol>
<li id="mpJdkt">Oklahoma (58.8%)</li>
<li id="Vjs3tN">Oklahoma State (58.1%)</li>
<li id="uhWtqd">Texas (54.1%)</li>
<li id="Je5Etr">Kansas (50%)</li>
<li id="Mz7ET3">Texas Tech (47.5%)</li>
</ol>
<p id="G1nYi6">Top 5 Success Rate, Third and Long</p>
<ol>
<li id="4Et08R">Oklahoma (31.3%)</li>
<li id="kdIXt1">Oklahoma State (29.6%)</li>
<li id="nFc1Bu">Baylor (26.9%)</li>
<li id="XLDyup">Kansas State (25.6%)</li>
<li id="5qs3TQ">Texas Tech (25.9%)</li>
</ol>
<p id="EJu22B">Third and medium is a mystery down of sorts - there isn’t an obvious trend in whether you should run or pass, and so the better offenses come to the top here (remember, we are proxying Kansas with Florida and Texas Tech with Utah State, the programs most recently connected to each team’s offensive manager). Aside from Oklahoma State, who had an incredible rushing attack combined with the WR talent to make defenses indecisive in their third down coverage, all four teams have elite QB play. The quarterback matters on third down more than anything, and on third and medium, elite QB play has a time to shine, the defense on their heels. </p>
<p id="xN0FBK">Third and long could be conceived as a predictor of some intangibles - poise, grit, confidence, etc. - we see two teams creep up onto this list, and Kansas State is really an oddity due to sample size. Baylor, on the other hand, has a dual threat QB who learned a lot last season and is regarded by many as a fierce competitor. This season, plenty of signs are pointing towards Baylor having a better offense than they’ve had in years. I’d expect Baylor’s offense to be similar to the mold of Texas and Ehlinger this season, based on small indicators like this. </p>
<h1 id="RPcB3v"><strong>What About TCU? </strong></h1>
<p id="3TNyWH">The Frogs are effectively an outlier on both initial graphs of Big 12 Performance. TCU faced an average third down distance a quarter of a yard shorter than the second place team. In fact, TCU was 11th nationally in third down distance, which, under normal circumstances, would be a huge indicator that something was going right. The problem is, of course, TCU’s success rate. Looking at the Rushing vs Success Rate graph, we see a relatively balanced TCU, willing to run on third down. On third downs, balance is not a virtue. TCU’s third down struggles can be traced entirely back to quarterback play, and that “balance” is a result of play-calling without an identity, without a sure idea of what to do next. </p>
<p id="neohGc">As for the shortest average distance, it’s worthwhile to discuss a theory as to why that, too, is misleading in terms of TCU’s offensive quality. TCU ranked in the bottom third of FCS offenses last year, largely due to injury and inconsistency. No one is here to throw stones at the Mule, a man who earns every ounce of praise he gets. Instead, I’m offering a simple theory: TCU faced the shortest average third downs in the Big 12 because most other Big 12 offenses converted their second-and-manageable downs. TCU was 78th nationally in “First Downs Coming on First and Second Down” at 67.5%. (For reference, Oklahoma was first with 80%.) So, TCU’s apparent “success” in getting to manageable third downs is really more of a failure. </p>
<p id="oAyUtq">On third and short, TCU’s success rate was only 62.5%. They rushed 73.3% of the time, which proved to be a strategical advantage - their rushing success rate on third and short was 20 percentage points higher than their passing success rate. Still, though, TCU ranked 7th in the conference in third and short success rate. </p>
<p id="jYnPFi">On third and medium, TCU’s lack of offensive identity became more apparent. TCU’s success rate fell to 9th in the conference - 33.9% - as their rush rate fell to 20%. For whatever reason, the Frogs felt like they couldn’t rush and succeed on short and medium; this might be related to offensive line, or more realistically just reveal a preference for situational passing. Third and medium is the area I’d keep an eye on most this fall, especially that rush rate. With the two-headed monster of Darius Anderson and Sewo Olonilua (we hope!), and a couple of more than capable rushers at QB, the rushing attack should take some precedence and have a little bit bigger of a role for TCU this year. </p>
<p id="edoEQa">Finally, on third and long, TCU rarely converted. Given their short average distance, they very rarely saw a true third and “long”, but where they did, only 15% of the time did they convert, Lending credence to my argument above, TCU’s rush rate stayed steady at about 20% on third and long. </p>
<p id="E5bj6R">TCU’s third downs last year displayed their troubled search for an identity amidst some horrible injury luck and some unmet expectations. The most successful teams on third down were teams with established QB play, and TCU will look to a deep QB room to find someone to come in, take over, and keep drives going this fall. </p>
https://www.frogsowar.com/2019/7/19/20698112/stats-o-war-season-preview-third-down-offensesstatsowar2019-07-11T12:30:00-05:002019-07-11T12:30:00-05:00Stats O’ War Season Preview: Second Down Offenses
<figure>
<img alt="" src="https://cdn.vox-cdn.com/thumbor/cO704QrcNjY5KqIyPI4YYun54Bk=/0x119:500x452/1310x873/cdn.vox-cdn.com/uploads/chorus_image/image/64709767/SOW_2019.0.jpg" />
</figure>
<p>The Stats O’ War preview is back, and this week’s installment breaks down team tendencies and performance on second downs across the Big 12.</p> <p id="xF5Y7H"><em>Author’s note: This is the second in a three-part offensive scouting exercise in preparation for the 2019 season, and the second entry in Stats O’ War’s inaugural College Football Season Preview. You can find the First Down Scouting Preview </em><a href="https://www.frogsowar.com/2019/6/27/18693180/big-12-preview-offenses-first-down-advanced-stats"><em>here.</em></a></p>
<p id="sKNbxS"><em>The first part of this post focuses on TCU’s Big 12 opponents. If you’re just here for info on the Frogs, you can skip down to the “What About TCU?” section at the bottom.</em></p>
<hr class="p-entry-hr" id="4qhDTz">
<p id="yhCODP">Whereas first down is a chance for an offense to establish itself, to set up its preferred play sequence, or put an opponent on his heels, second down is more boom and bust; if an offense takes care of business on first down, then second down can be a chance to shoot your shot, a free play. If an offense fails to set itself up for success on first down, second down becomes a mad scramble for yardage. </p>
<p id="S245xQ">In the high-speed spread world of the Big 12, second down is where teams can twist the knife, make up ground, or catch teams off guard. Last time, I looked at tendencies of offenses on first downs in different situations, specifically focusing on the first play of a possession. I found some shocking behavioral differences among teams depending on their field position and on the first down of a series. This week, I’m diving into second downs to continue my stats-based Big 12 Offense Scouting Preview. Below, after a quick word on methodology, I’ll lay out and comment on some descriptive statistics of second downs, examine rushing and passing tendencies for each team, and then discuss the context for big plays. </p>
<h1 id="pUivRj">A Word About Methodology</h1>
<p id="IHTyoD">If you’ll remember, loyal reader, in my first downs preview I mentioned I had done something a little weird in preparation for the 2019 season. There’s been a lot of turnover in the Big 12, and so to most consistently align expectations with what we will see on the field, I threw out data of three actual 2018 Big 12 teams, and replaced it with proxies for their new coaches. So, below, when you read “Kansas”, “Texas Tech”, or “West Virginia”, that really means “the offenses most recently associated with the coaching staffs headed to those three schools. As for the non-FBS coaches (Kansas State, Oklahoma State’s OC), I left those alone as I figured their style wouldn’t change too dramatically this year, but also because I don’t have the D1 data. Now, on to the fun stuff. </p>
<h1 id="n4hOGR">Second Downs: Descriptives</h1>
<h3 id="Ju91FN">Yards Per Play on 2nd Down: </h3>
<ol>
<li id="R6fimi">Oklahoma 8.52</li>
<li id="6yFh0E">Texas Tech 7.76</li>
<li id="zgRUjA">Baylor 6.17</li>
<li id="irWtFN">Oklahoma State 6.07</li>
<li id="o4AIvH">TCU 5.83</li>
<li id="Fgu4sR">Iowa State 5.81</li>
<li id="nlTtKw">Kansas 5.54</li>
<li id="BxYqRv">Kansas State 5.51</li>
<li id="C9cObS">West Virginia 5.45</li>
<li id="P4l8iv">Texas 5.41</li>
</ol>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/CKNdS-oyN7MiyHSlq3BPo47vf_s=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/18302852/2downypp.jpg">
</figure>
<p id="TluYAk">No surprises here - OU was the most dominant offense in the Big 12; if the Sooners didn’t get a first down on first down, they definitely made up for it on second down. The field flattens out after an explosive Texas Tech - Matt Wells’ offense was a pain in the neck of most Mountain West Defenses last year. Coming into his first season in the Big 12, that shouldn’t be too much the case - Wells’s offense was driven by his developed upperclassmen roster, and it’ll take some time before Lubbock gets that pointsy. The Kansases, along with Iowa State, West Virginia and Texas round out the bottom of the list - Iowa State and Texas committed to their chunk-of-yards consistency, WVU featuring a spread option, and and the Kansases, well, they do what one does in Kansas. </p>
<p id="nWFiKp">For all but two teams, (the Kansases), yards per play on a pass were much higher than yards per play on a rush. That’s to be expected, generally, but also indicates another pattern - teams in the Big 12 are rushing to make up yardage on second down, passing to take a shot downfield. (Now, I don’t have data on passing attempts, but we can infer this situationally.) OU, Baylor, TCU, and West Virginia seem to have the biggest splits, and we’ll dive into that deeper below. </p>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/vAV0M9_2q1uWOHTf_4NpDt3mK_w=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/18302853/2nddownsuccess.jpg">
</figure>
<p id="hqQk40">Most teams are fairly balanced, with a few outliers, in terms of relative success of rushing and passing. West Virginia (Neal Brown’s “not an air raid” offense) was much more successful on the pass, as was Texas Tech (Matt Wells’s “not an air raid” offense). One could argue that Kansas and Baylor were more successful on the run than the pass on second down, but just barely. If we were to discuss this in terms of optimization, then, perhaps we may presume that most teams found a proper mix as to equalize the comparative advantage of rushing and passing. In theory, a team who is drastically more successful at one over the other should strictly benefit from producing where they have the comparative advantage. The only consideration is that of strategy - if WVU is only good at passing on second down, and they rush very few times, then their passing game will be diminished by a defense who has an idea of what to expect. So, we’ve seen these raw numbers on success and yards per play, but let’s look at strategy to see who’s doing what:</p>
<h3 id="oekEqT">Second Down Rush Rates:</h3>
<ol>
<li id="6v3dP4">Kansas 61%</li>
<li id="jEOj2c">Kansas State 60%</li>
<li id="6lAHA1">Texas 59.3%</li>
<li id="MfzgQz">West Virginia 57.1%</li>
<li id="cgEvfL">TCU 56.9 %</li>
<li id="6URcA7">Iowa State 55.3%</li>
<li id="jEnV5z">Baylor 50.9%</li>
<li id="81IHVr">OU 50.6%</li>
<li id="NY3wj2">Texas Tech 49.8%</li>
<li id="GbKE4N">Oklahoma State 48.5%</li>
</ol>
<p id="KT0EoU">Kansas, Kansas State, Baylor, and Oklahoma State are spread across the range of rush rates, all with similar comparative success - (run vs pass). This seems to bolster the argument that coaches are optimizing to try and equalize their comparative advantage. Kansas and Kansas State had bad passing games, so they rush more in order to compensate for quality. On the other hand, Oklahoma State committed to the pass, rushing just enough on second down to keep opponents honest. WVU actually runs more often than most, which brings up a stark contrast with Texas Tech - both teams have a drastic comparative advantage in passing, and Texas Tech steers into it, while WVU sat on the run. Something to look for this season will be how Matt Wells and Neal Brown adjust their play-calling relative to comparative advantage.</p>
<p id="j0DJ16">The top of the list features the four lowest yards per play second down offenses, highlighting perhaps how some teams run on second down out of principle. For instance, multiple times last year we lamented how TCU would often run up to the line on second down and run a throwaway dive play, effectively wasting a successful play. That commitment to the run seems to have inherently limited the ceiling of those top six, who as we will see below, tend to have fewer big plays. </p>
<h1 id="oaknKO">Contextual Second Down Statistics</h1>
<p id="f8yEQT">The raw numbers help us paint a picture of which teams succeed where on second down, but second down has a lot of contextual factors. Rather than focus on short, medium, and long, I want to condition plays based on the play before. What I’ve done is categorized plays as occurring after a run, after a pass, after a success, and after a failure. That should capture much of the variation in distance, while providing more information about team tendencies overall.</p>
<h3 id="NfQ5py">Most Resilient 2nd Down Teams (Success Rate after Failure):</h3>
<p id="g9L82h">I’ll start with a stat I’m calling resilience: Which teams were able to find success on second down after a first down failure? </p>
<ol>
<li id="GolUeI">Baylor 34.6%</li>
<li id="ozw3G6">Oklahoma 34.1%</li>
<li id="3Q6UTd">Texas Tech 32.8%</li>
<li id="sK6jyF">Kansas 31.6%</li>
</ol>
<p id="0N6FIt">These four top the list (the rest of the data is in the table below), and tell a couple distinct stories. Oklahoma, of course, has an incredible offense, and just didn’t fail twice that often, so we’d expect them to be up there in terms of bouncing back. Texas Tech, under Matt Wells, featured an experienced and talented roster, so the same argument applies. Baylor, though, was the inverse: a younger team, struggling to find some identity, had its positive moments. This stat might indicate, coupled with returning production numbers, that Baylor’s offense should look much better this coming season - they are resilient and experienced, a dangerous combination. </p>
<h3 id="yBe30l">Most Consistent 2nd Down Teams (Success Rate after Success):</h3>
<ol>
<li id="kOgc4A">Oklahoma 17.5%</li>
<li id="2TIJxC">Texas 15.3%</li>
<li id="2yLDbU">Iowa State 13.8%</li>
<li id="R8f2AZ">Texas Tech 11.5%</li>
</ol>
<p id="KM2eYY">Here again, we see Oklahoma’s dominance, quantified. Iowa State and Texas, with their low YPP on second down, both sit at the top of this list. This is in line with their consistent approach; although neither team ranks near the top in yards per play on any given down, they are consistently gaining, playing three down ball to get first downs most of the time. It appears that under Matt Wells, Texas Tech will fall into more of that “slow and steady” approach - consistent low ceiling success, resilient when you fail.</p>
<div id="JBmtb5"><div data-anthem-component="table:8656281"></div></div>
<h1 id="fTPo7A">Second Down Strategy</h1>
<p id="hAETEE">The next question, after success and failure analysis, is one of strategic considerations. We’ve seen above that some teams have a comparative advantage, and some teams are balanced in run and pass - how did they get there? Most teams favored a pretty equitable spread in tendencies, ranging in the mid-to low twenties for each of the four possible combinations. Discussing the outliers will empirically confirm some tendency assumptions we already have.</p>
<p id="350D8w">Kansas and Kansas State, due to talent or due to style preference, have the highest rates of rush plays after a run. Texas and TCU both have abnormally high run rates after a run - Texas more so to keep the momentum of the QB going, setting up for the big play. Oklahoma State, on the other hand, keeps up their reputation with the second lowest rush-rush percentage. The Pokes bring in a creative an unorthodox offensive coordinator. It’ll be of note to see how he handles OSU’s commitment to passing early and often in a set of downs. </p>
<p id="w92Qz8">Baylor leads the league in the “At Least One Pass” category, meaning that the Bears were going to pass on either first or second down most of the time (78%). Some work on conditional probability tells us that if Baylor didn’t pass on first down, that 50% rush rate from above would actually increase to the 75% range - that’s a key tendency to exploit, and seems like one of the few obvious notable habits in this data set. <br><br>Kansas State and Texas lead the league in the “At Least One Rush” category; Kansas State’s devout commitment to the run explains their tendency there, nothing new. Texas, as we saw above, really prefers to chip away at yards on first and second down, setting up a manageable third down, and so their tendency isn’t exactly an astounding revelation. The “At Least One” categories become another fixed point for us to watch with the new coaches coming into the Big 12 this year. </p>
<h3 id="qvPmsr"></h3>
<div id="f97Ohb"><div data-anthem-component="table:8656275"></div></div>
<h1 id="B3EW7v">Second Down Big Plays</h1>
<p id="tgQjHX">Above, I’ve noted that second down can be a chance for an offense to “take a shot” after a successful first down, to try and capitalize on your talent with the safety net of a manageable third down in tow. How successful were offenses at taking shots on second down? Who did the most damage? To find that out, we’ll look at big plays. Big plays, according to my arbitrary definition, are any rushes > 10 yards and any passes > 15 yards. These aren’t just your big-breakaway touchdowns; these are meaningful chunks of yards gained. </p>
<p id="uMy3qG">The more pass heavy offenses lead the league in big play percent, but I want to focus on two specific oddities. Iowa State and Texas. Both teams are of the “consistent low ceiling high floor” style when you look at yards per play and success rate. Big Play rates give us a different story. Iowa State comes in third in Big Play rate, behind the dominant Oklahoma and the pass-happy Oklahoma State, but leads the league by 4 percentage points in “Big Plays after Success.” <strong>That is to say, Iowa State gets more of their big plays after successful first downs than anyone else in the Big 12. </strong></p>
<p id="sNdMXp">What a credit to Matt Campbell! The Cyclones don’t just stick their head in the mud and commit to gaining chunks of yards - when they see an opening, they take it, and they take advantage. This is also a huge credit to having a high caliber running back (most of their big plays were rushes) who can sift through a defense looking to stop something on second-and-manageable. </p>
<p id="guUtZ1">Texas, on the other hand, confirms their commitment to playing three down ball with the lowest second down big play rate in the league. The Longhorns aren’t going to come after you with gambles, they’re going to make you defend nose-to-nose over the course of three downs, hoping to wear you out. That subtle shift in play-calling can counter-balance some of the write-offs of Texas, that they were “lucky” to win close games, that they got a lot of “help” to win those games. Those write-offs are true, but are somewhat mitigated by an intentional preference to wear down an opponent and then take advantage of them. With the lowest returning production among Power Five schools next year, Texas will definitely see a set-back, but expect Tom Herman to take every advantage of his talent with his “make-you-defend” style.</p>
<div id="TRu9I9"><div data-anthem-component="table:8656325"></div></div>
<hr class="p-entry-hr" id="S0kaIg">
<h1 id="hVjw6z"><strong>What about TCU?</strong></h1>
<p id="TYqacE">TCU is in the top half of resilience, succeeding after a failure 30.1% of the time. Some of that we can attribute to game plan - first down screens aren’t often successes, and a strong run game can push second down into a manageable third down (which is the goal, according to our definitions of success). The trouble for TCU is that more than <sup>2</sup>⁄<sub>3</sub> of the time, if they failed on first down, they failed on second down. In fact, TCU followed up a successful play with a successful play only 8% of the time last season. That does not bode well for mounting long, successful, scoring drives. TCU more than half of their possessions (52%) failed on both first and second downs. That effectively puts TCU “in-the-hole” on third down 92% of the time! (You’ll recall, TCU faced one of the longest average 3rd down distances last year.)</p>
<p id="vRKrLn">Offensive line play should improve (<a href="https://www.frogsowar.com/2019/7/10/20688364/tcu-horned-frogs-football-offensive-line-preview-lucas-niang-anthony-mckinney">see Grant’s preview of the Big Guys her</a>e), and the playmakers are there (<a href="https://www.frogsowar.com/2019/6/26/18716005/catch-me-if-you-can-previewing-tcus-wide-receivers">Melissa’s WR preview here</a>), but the quarterback is a question mark. The relative instability of TCU’s success, plus the absolute instability of the quarterback situation gives us an easy barometer to watch. How will TCU approach second down going forward, knowing that in the past, they’ve struggled to string together multiple successes?</p>
<p id="GUldnU">The Frogs find themselves towards the top of the rush-rush percentage, and I believe that is even biased downward - recall, the Frogs’ offense of 2018 was horizontal, and screens to the flats on first down are effectively rush plays, categorized as passes. Additionally, TCU ranked fourth in pass-pass rate; that, I believe, is due again to their swing/screen pass focus last season. </p>
<p id="s9V5p3">As for mixing their plays, the Frogs sit slightly above average on the “At Least One Rush” category, and of course, slightly below on the “At Least One Pass” category. I can’t help but mention the downward bias in the rush rate, and so you’ll see that last year, TCU actually preferred to run pretty heavily. TCU is going to run on first or second down almost 80% of the time, so that 56% second down rush rate from above actually becomes much higher when TCU doesn’t rush on first down. </p>
<p id="6sUNe7">The Frogs ranked near the bottom in Second Down Big Play percent, at 13.4%, characteristic of a stagnant offense. They found more explosiveness with the pass than the run (thank you, Mr. Reagor), as more than half of their Second Down Big Plays came from passes. The Frogs, as indicated above, struggled to string successes together, so of course they had the fewest big plays after successes of anyone in the league. Neither did the Frogs show any tendency to follow up a rush or a pass with a big play - TCU was evenly split on big plays after rushes or passes. </p>
<p id="uoAZJ7">There have been rumors that TCU is going “vertical” this offseason - will that result in fewer rush-rush combinations and more big plays as the coordinators focus on the run as a device to set up the long ball? The talent is there for TCU to have a 2014-style offense, on paper, but the Frogs will have to adjust their play-calling tendencies and execute in order to reach their ceiling. </p>
<p id="wXB0Mq"></p>
https://www.frogsowar.com/2019/7/11/20689566/stats-o-war-season-preview-second-down-offensesstatsowar2019-06-27T11:30:00-05:002019-06-27T11:30:00-05:00Offenses, Pt 1: First Down
<figure>
<img alt="Baylor v Texas Tech" src="https://cdn.vox-cdn.com/thumbor/sCm6cFC7O-FLkDNc1b4UoJR4G-A=/0x0:2927x1951/1310x873/cdn.vox-cdn.com/uploads/chorus_image/image/64141682/625825158.jpg.0.jpg" />
<figcaption>Photo by John Weast/Getty Images</figcaption>
</figure>
<p>More than you ever thought you’d want to know about how Big 12 Offenses approach First Down</p> <p id="DOrO7x"><small><em>Author’s note: This is the first in a three-part offensive scouting exercise in preparation for the 2019 season, and the first entry in Stats O’ War’s inaugural College Football Season Preview. To start, I will analyze Big 12 offenses, down by down, to see what some more obscure numbers reveal about the style, approach, and attacks of each TCU opponent.</em></small></p>
<hr class="p-entry-hr" id="xQZeL0">
<blockquote><p id="dSOCCC"><em>The secret to getting ahead is getting started.</em> - <strong>M. Twain</strong></p></blockquote>
<p id="3HL8zO">Each possession, each fresh set of downs provides an opportunity for an offense to start over, to try something new, to approach the problem of gaining yards in a new and better way. A successful play on first down can set the tone for a drive, getting an offense proverbially “ahead of the chains,” whereas a hole dug on first down by a failed play lingers. </p>
<p id="T86e4D">Aggregate statistics - run percentage, pass percentage, and success rate, for example - constitute a baseline analysis of an offensive style, but lack the nuance of context. One can begin to grasp the fundamental approach of an offense via a thorough analysis of contextual first down behavior. In this article, I will outline some descriptive statistics concerning the tendencies and preferences of Big 12 Offenses (and some newcomers!) on first downs last season, examine the ways in which those patterns changed when the first down was the first play of a possession, and discuss how these situational adaptations inform our expectations of TCU opponents in the coming season. </p>
<p id="RRCwHm">First, a brief segue to discuss some data and methodology. </p>
<h1 id="4RZqGy"><strong>Coordinators, Old and New</strong></h1>
<p id="yjnmBZ">As Big 12 football fans - both those astute and aloof - are well aware, there are some new coaches faces in the Big 12 this year. That context of course necessitates some adjustment to the data I consider in the ensuing analysis. </p>
<p id="c9Na8Q">As Texas Tech and West Virginia brought in coaches and coordinators from FBS schools, I can easily add in Utah State (Matt Wells at Tech) and Troy (Neal Brown at WVU) to understand the types of offenses these new coaches might run. </p>
<p id="PTUoDA">Two coordinator situations offer no direct parallel - Chris Kliemann brought his NDSU partner in crime Courtney Messingham to Kansas State, and Mike Gundy at Oklahoma State hired Princeton’s Sean Gleeson. In both cases, I do not currently have access to granular data at the FCS level, so I am limited in how thoroughly I can cover each coordinator’s past. Instead, I’ll argue that at least for this year, Kansas State should still be run-heavy in style and Oklahoma State isn’t changing the scope of their high-tempo offense, just the presentation. Therefore, in both cases, I’ll stick to the Big 12 teams’ 2018 body of work for analysis. </p>
<p id="Oge7FV">Finally, there are two more changes in terms of coordinators: Iowa State brought back Tom Manning from the NFL, and Les Miles hired Les Koenning at Kansas (Les to the power of two!). I’ll keep Iowa State’s analysis restricted to 2018, as Manning had previously been at Iowa State, and so I don’t expect a huge swing in style this coming season. As for Les Koenning, I’m going to do something a little weird - Koenning’s last meaningful job was under Dan Mullen at Mississippi State, so in lieu of Kansas’s 2018 data, I examine the tendencies of Dan Mullen’s Florida offense to approximate what we may see in a best-case-scenario for Kansas. </p>
<p id="JLPsLm">To clarify, going forward, I’ll be discussing only the Big 12 teams, but when I say “West Virginia” I’m really looking at Troy’s data, and the same for “Texas Tech” and Utah State, and “Kansas” and Florida. I hope that’s not too confusing. Anyway, on to the data. </p>
<h1 id="pGHAAt"><strong>First Downs - Descriptive Statistics</strong></h1>
<p id="jkoadv"><strong>Top 10 First Down Offenses, by YPP (2018):</strong></p>
<ol>
<li id="Aa6iHs">Oklahoma 8.6 ypp</li>
<li id="LHkiWp">Iowa State 6.9</li>
<li id="v7RxD1">West Virginia 6.6</li>
<li id="9xNjot">Baylor 6.2</li>
<li id="VYPt3H">Oklahoma State 6.2</li>
<li id="tZjvY1">Kansas 6.1</li>
<li id="B9iuMe">Texas Tech 6.0</li>
<li id="RZhklU">Texas 5.3</li>
<li id="UQGePm">TCU 5.2</li>
<li id="wZoMDk">Kansas State 5</li>
</ol>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/1BkDNK_5gB5oNl4k5lj_lbp0DRA=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/16446693/First_Down_YPP.jpg">
</figure>
<p id="2l002i">A bar chart provides a simple level comparison of each team’s yards per play overall, and for rushes and passes, on first down. The dotted line represents the average YPP for all first downs for all ten teams, 6.22. There is little variation, for the most part, as most teams’ offenses are above 4 yards per play and all but two passing units are under about 8 yards per play. Oklahoma, of course, stands alone as having all three units above average. </p>
<p id="MTYHfv">This graph also highlights the variety of styles in the Big 12, generally. Some teams are consistent big gainers on first down, and some are small gainers, but neither seems tied particularly to success - Texas, for instance, is near the bottom in total YPP on first downs, but still had a relatively successful offense last year. </p>
<p id="hSjbyb">We can categorize some of the asymmetry to paint in broad strokes about style. Pass-success teams include: Baylor, Kansas, Iowa State, Oklahoma, Oklahoma State, Texas, and West Virginia. Interestingly, no team gained more on rushes, but that’s going to be weighted and skewed by the fact that passes are necessarily longer attempts most of the time. An offense like TCU, for example, that has historically been screen-heavy, has a flatter distribution between success in the run and the pass. </p>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/i51LWJSJ72J7NqL9drT0gz6wIYg=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/16446717/FD_PassRush_Success.jpg">
</figure>
<p id="HHTGsz">Plotting teams’ rushing and passing success rates on first down further clarifies our initial inference; most teams skew pass heavy, in terms of success on first downs (anything right of the dotted red line indicates more passing success). Again, Oklahoma stands out, as they do in all offensive categories, but the clustering in the center reveals some stylistic successes. Oklahoma State relies on the run much more than their speed spread reputation suggests; Matt Wells has roots in the Air Raid, but is decidedly not an “Air Raid” guy, and we see a bit more balance. Similarly, we see Baylor, TCU, Kansas State, and Troy (Neal Brown moving to WVU) all rely more on short yardage first downs, again reflecting the diversity of style and execution - TCU heavy on screens, Troy a run team satisfying WVU’s urge to move away from the Holgo-Air-Raid, and Kansas State’s commitment to bread and butter, three yards and a cloud of dust. Initially, first down success seems to be a mixed bag - the ideal of five yards each first down is certainly desirable, but by no means totally indicative of whether each play was a “success”. </p>
<p id="94aDem">We’ve seen success, and who makes their money on the run and the pass, but how does strategy parallel or drive that? Here are the first down rush rates for each team: </p>
<ul>
<li id="ZRZSwa">Kansas State 67.7%</li>
<li id="zY7nEz">OU 61.3%</li>
<li id="s0ZZAl">TCU 60.1 %</li>
<li id="WQm0f9">Kansas 59.9%</li>
<li id="0OixUC">Iowa State 58.9%</li>
<li id="9yNCVC">Texas 54.0%</li>
<li id="SxizMC">WVU 52.2%</li>
<li id="9odnlg">Baylor 50.9%</li>
<li id="UMvTyf">Oklahoma State 50%</li>
<li id="yRvgUL">Texas Tech 47.7%</li>
</ul>
<p id="pUTHph">Kansas State, the least successful rushing team, is also the team who rushes the most on first downs. This reflects that style commitment to establishing the run. For the Wildcats, evidently the number of yards matters less than the number of opportunities they have to be physical with their opponents (the body-blow theory). </p>
<p id="TWHZW7">TCU’s 60% is perhaps deflated on first downs, as the Frogs like to spread their run game outside with short screens. Whereas teams like Oklahoma State, Baylor, and Texas Tech prefer to work in longer shots on first down, the Frogs focus remains on consistent first down yardage. Will that change with the rumored return to the downfield veer and a more “vertical attack”? </p>
<p id="yXSrr5">The Big 12 has a reputation for pass-happy, high tempo offenses. Below the surface, though, establishing the run is paramount in offensive success in the Big 12 - 8 of 10 teams rush more than half the time on first down, and half the league rushes more than 57% of the time on first down. </p>
<p id="6oxLjg">This first down data gives us an idea of how each team attacks and succeeds on first down, but the aggregate numbers obscure team intent - the context of each first down and the plays run leading up to each first down dictate much of a team’s approach. To adequately comprehend each team’s preferences, let’s examine how they behave on first and possession - the first first down of each drive. </p>
<h1 id="axPVrK"><strong>First and Possession: Trends and Differences</strong></h1>
<p id="SMUS77">Why look at first and possession? While this eliminates some of our sample size, isolating the initial plays of each drive strips away most of the confounding context and reveals to us each team’s preference about how to approach a drive. Granted, there are still some game state realities that dictate strategy, but getting that granular really inhibits the meaningful comparisons we can make, in terms of sample size. For now, let’s just stick to a broad analysis of success rates and rush rates, comparing those to all first downs, and see what we learn.</p>
<h3 id="Lbn9ru">Success on First and Possession: </h3>
<p id="ajy6aS">Top 10 Offenses on First and Possession, by YPP</p>
<ol>
<li id="hnzCKK">Oklahoma 8.6</li>
<li id="gb8MMS">West Virginia 8.2</li>
<li id="z4NFZS">Iowa State 6.8</li>
<li id="f5pPzg">Baylor 6.7</li>
<li id="U5iewN">Texas Tech 6.6</li>
<li id="e9PfCr">Kansas 6.5</li>
<li id="hpfzet">Oklahoma State 6.4</li>
<li id="3xxDkO">Kansas State 6.2</li>
<li id="gEmInP">Texas 5.7</li>
<li id="IL4JKd">TCU 5.3</li>
</ol>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/Bfh9tDfIRFaH2lgS8FUHRUVaQB0=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/16677306/first_pos_ypp.jpg">
</figure>
<p id="PASAxS">Starting out with raw yards per play, I plotted the average gain on the first play of a possession against the the average gain on all first downs. We see some stark differences here. All 10 teams sit above the 45 degree line (the black line, indicating the point where First Down YPP = First and Possession YPP): teams gain more on the first play of a possession, on average, than they do overall. </p>
<p id="sD5s0i">Consistency, however, doesn’t seem to factor into total YPP - the top four teams in YPP on first and possession are split between equal success on first plays of possession and first downs (ISU, OU) and being drastically more successful on first plays (Baylor, WVU). We see at the bottom two teams who struggled for different reasons - TCU’s offense was a mess at times, and digging a hole on first down didn’t help. Texas, on the other hand, is the odd case of consistency - they weren’t very explosive on first down, but they were consistent. </p>
<p id="T4tX8f">Kansas State and West Virginia are the two biggest differences between first and possession and first downs: West Virginia (Troy) through taking shots downfield and Kansas State due to their plodding run commitment. </p>
<p id="5cKgZk">The above scatter plot lends credence to the utility of examining first and possession, though, as we see teams perform better when they presumably have the strategic advantage of the fresh slate of a drive. </p>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/-zfDdaWh4Srr47OHflmtFabvst4=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/16677315/first_pos_sr.jpg">
</figure>
<p id="yKAnDi">Looking at success rates, we note immediately that Oklahoma was, in layman’s terms, a damn good offense last year. They didn’t beat you situationally, they just beat you. Also of note, their difference in success rate comes from the fact that they scored early and often, and their drives are notoriously short due to their explosive offense. </p>
<p id="DXKBrr">As for the rest of the conference, we see Texas Tech, Iowa State, Texas, and TCU all less successful on first and possession than first downs, and again we see a variety of drivers: Texas Tech, as we will see below, passed the most of anyone on first and possession, which trends toward a lower success rate. Iowa State and Texas modeled more of that consistency, opting for short gains rather than long strikes, especially on the first play of a drive. </p>
<p id="jw3iSt">In the case of TCU, we see yet again what it looks like to stall out as an offense. TCU came out of the gate slow on most drives, unable to establish an identity on first downs, which obviously bled over into the resulting second and third and longs. </p>
<p id="mQ4dTJ">Oklahoma State, Baylor, and Kansas all came out as the “feel-good” offensive teams of the year - when the drive started, these teams were able to put together some yards, but that ability waned as the drives continued. All three were talented teams with serious flaws, but when they were at their best, they could put together yards and establish the start of the drive. </p>
<h2 id="ZEQRqM"><strong>Run/Pass Tendency and Success</strong></h2>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/vEzVGJM5GBee9sdrCQW-0Jn8AyY=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/16677333/ypp_fd_fp_rush_pass.jpg">
</figure>
<p id="NcJCM1">Adding in the YPP numbers from first and possession to the bar graph from above, we see some spectacular consistency across run and pass abilities. The only significant discrepancies are Kansas State’s pass game, and arguably, Baylor’s pass game. On the whole, the levels shift down, on average, slightly, but for the most part the relative success of rush and pass remains unchanged for each team. </p>
<figure class="e-image">
<img alt=" " data-mask-text="false" src="https://cdn.vox-cdn.com/thumbor/eUQpmMbokPM0TGw-sjeA4MSTKr8=/400x0/filters:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/16677336/fp_rush_rte.jpg">
</figure>
<p id="JAz1Bc">Finally, we come to rush rates. This is the nitty gritty of strategy, the true indicator of how a team establishes themselves on first and possession. Again, we see no large differences, but a few meaningful ones that might just inform us as to how teams in the Big 12 like to start their drives: </p>
<p id="6aKI9v"><strong>RUN-HEAVY: </strong>Kansas State, Iowa State, Oklahoma, Kansas, TCU<br><strong>BALANCED: </strong>Oklahoma State, Baylor, WVU, Texas<br><strong>PASS-HEAVY: </strong>Texas Tech</p>
<p id="zckz9n">For not being an Air Raid Guy(tm), Matt Wells’s offenses sure do like to pass a bunch. </p>
<p id="xQWH20"><strong>RUN MORE on F&POSS: </strong>Oklahoma State, Baylor, OU, Kansas, Kansas State<br><strong>PASS MORE on F&POSS: </strong>Iowa State, TCU, Texas, Tech, WVU</p>
<p id="bwTtcn">The league is almost split 50-50 on how they approach the game. TCU, as I’ve mentioned before, has some deflated rush numbers because a lot of their offense is really a screen run game, so you could almost put them in that first category. Despite the Big 12’s reputation for slinging the ball, most teams prefer to establish the run regularly on first down. </p>
<h1 id="o7x4Tg"><strong>Conclusion</strong></h1>
<p id="8sGBHk">The main conclusion to be drawn from this analysis on first downs is that, despite what the memes and the talking heads would have you believe, Big 12 offenses actually establish the rush in their game plans. Big 12 teams rushed over 56% of the time on the first play of a possession. The data indicates that on the first play of a possession, the league did split tendencies: Oklahoma State, Baylor, Oklahoma, Kansas, and Kansas State all preferred to run more to start out drives, whereas Iowa State, TCU, Texas, Texas Tech and West Virginia all committed more to the pass to start out their drives.</p>
<p id="4yW3bv">The novel contribution of this article is that by examining the offenses of Utah State (Texas Tech), Troy (West Virginia), and Kansas (sort of Florida) in context, we were able to predict the best case scenario for three future TCU opponents and see how they fit into the Big 12 landscape. (WVU might pass less, Kansas will be more balanced, and... Tech’s going to still do the Tech thing. In fact, Tech might be even more Tech with the addition of Matt Wells.)</p>
<p id="ppL9cA">All in all, this article provides a solid starting point for a thorough analysis of Big 12 offenses, and lays a foundation of discussion moving forward. It’s important to note that first down is only one piece of the puzzle, and even that breaking tendencies up by down is only one piece of the puzzle, but this piece of the puzzle gives us plenty of insight into what we can reasonably expect from Big 12 offenses going forward. </p>
<p id="1pxFeg"><em>This is part one of many in the Stats O’ War 2019 Preview Series. You can follow Stats O’ War on Twitter (@statsowar) for more content and updates about future installments. </em></p>
https://www.frogsowar.com/2019/6/27/18693180/big-12-preview-offenses-first-down-advanced-statsstatsowar