What are the ‘Noles getting in Norvell?

FSU just hired Memphis coach Mike Norvell to take over the Noles football team. As I did prior to UF’s hiring of Dan Mullen and FSU’s acquisition of Willie Taggart, I took a look at how the coach had performed at his prior post in terms of winning and recruiting. Here is a quick and dirty look of what I found on Norvell.

On-Field Performance

Memphis was on the rise prior to Norvell taking over. This often makes sense at smaller schools- they are often replacing successful coaches when those coaches get offers to bigger programs, as was the case when Norvell took over for Justin Fuente, who left to coach Virginia Tech.

norvell charts

As the above charts show, Memphis was performing fairly well prior to Norvell. To his credit, they didn’t fall much once he took over. The tables below outline the numbers.

tables w norvell

The columns with heat mapping are standardized values. The thing that stands out about Norvell is that his offenses have always performed above the conference average, even in a time in which UCF was lighting that league up. His offenses were 0.850, 1.573, 1.378, and 1.169 standard deviations above the league mean.

Recruiting

In looking at Memphis’ recruiting during Norvell’s tenure, I found he has done pretty well compared to his competitors. Not great, but not poor either. The chart below shows the 247 Composite recruiting scores for Memphis vs the AAC from 2016-2019. In these years, he was 0.633 (2016), 1.419 (2017), 0.004 (2018), and 1.110 (2019)  standard deviations above the conference recruiting score means. So he essentially had an ok year, a good year, and an average year, and another good year.

recChart

Here is a look at the tables with the recruiting scores for the AAC during that time:

recruiting1

recruiting2

All in all, FSU is likely getting a better coach than Taggart, but that isn’t really saying much. It is difficult to project how he will do at a larger program. I guess we shall see…

 

The SEC- They Just Have More. A Recruiting Comparison by Conference.

As we head toward the early signing period, teams are frantically chasing recruits. When looking at how each of these teams may be doing in their endeavors, the starkness of difference among recruiting success (according to the composite ratings) by conferences really stood out to me. I decided to chart each of the Power 5 conferences along 3 dimensions: Number of Blue-Chip recruits, Average Rating, and Count of Top 1000 recruits.

The number of Blue-Chip recruits is simply those who are 4- or 5-star recruits. The average rating is just that- how each team’s (and collectively each conference’s) recruits average out in terms of rating. Only recruits rated in the top 1000 (as of December 7, 2019) are included in the analysis. Additionally, the analysis includes all commits and Crystal Ball leans. If a high school recruit does not have a favorite team to sign with (lean), he would not be included. So, we shall see just how accurate the Crystal Ball predictors are after NSD.

Recruiting Race by Conference Event #1: Total Count and BCs

In the graph below, we can easily see that the SEC is outpacing the rest by a comfortable margin. They have 245 commits/leans out of the top 1000 players, and 131 of those are Blue-Chips.

RR by conf

Here are how the conferences pan out by percentages:

RR by conf percents

The SEC is not only getting, or likely getting the most of the top 1000 players, they are cruising toward getting the lion’s share of the Blue-Chips.

Recruiting Race by Conference Event #2: Total Count and Average Rating

Ok, so we can see that the SEC recruits are likely to be higher rated on average because they have the most BCs. That kinda foreshadowed who would be leading this event. Here is how it shapes out to date:

RR by Conf Avg Rating

Yea, the SEC is blowing the competition away in this category. SEC recruits average a .9015 rating with the next highest conference being the Big 10 with a .8856 average.

The Blue-Chip Ratio Visualization

The graph below shows how each team is doing in terms of Blue-Chip ratio (among the top 1000). This really highlights how well the top SEC teams recruit relative to the rest.

BCratio

The more blue showing, the higher the BC ratio is. The larger the mark is, the higher the count is. LSU is masking Clemson due to overlap and Alabama is the blank blue square to the right of Georgia (name got blanked out for some reason). The larger the square the higher the count.

Conference Breakdown

All conferences have teams that are pulling more than their share of the recruiting weight. However, in the SEC there is a bit less of this.

ACCBig10Big12Pac12SECteams

When adding up each of the scores and then ranking the conferences across all 3 dimensions, the SEC is the champ, so far.

finalmedals

Bonus:

When breaking down the scores and doing some statistical stuff, here is how each team ranks overall:

teamsranking

LSU’s great season appears to be continuing off the field as well (this was written before the SECCG). The SEC has 5 of the top 8 teams overall. Badass.

Kyle Pitts and the Mackey Award snub: A statistical look.

There seems to be little disagreement that Florida tight end Kyle Pitts should have been among the 8 finalists for the 2019 John Mackey Award, given to the nation’s top collegiate tight end. Statistically, Pitts clearly belonged in this group. Pitts’ stats are excellent and compare well with any of the top tight ends in the country. Here is a look at how Pitts’ key statistics would stack up to the actual finalists:

actual

Upon standardizing each of the three key categories under consideration (receptions, yards, and touchdowns) and totaling the standardized score, Pitts ranked 6th out of the (now) 9 players. This supports the popular notion that Pitts was snubbed. However, to look at how Pitts alone was snubbed wasn’t enough. I wanted to see how all of the tight ends would stack up to see if the committee got much right at all, or snubbed other players like they did Pitts.

Stats projected for 12 games

Using the top 50 tight ends according to https://www.cbssports.com/collegefootball/stats/playersort/NCAAF/TE

I set out to project how each player’s statistics would look over 12 games. This helps control for players that have played in fewer games. I removed players with fewer than 8 games and was left with a set of 48 players. Here is the 8 finalists and Pitts’ projected 12 game stats:

PittsplusMackeyTable

The Top 50*

Actually, the top 48. Anyways, what I found was interesting to me. Taking the top 8 of this list would produce a significantly different list than the one the committee came up with. As previously noted, I standardized each of the scores around a mean of zero (producing a z-score). This puts each of the 3 statistical categories in scale with each other. I then added these 3 z-scores to get an aggregate z-score (TOTz). The graph below shows how the top 48 would stack up:

Top50graph

The top 8 in this analysis still include Pitts (at #8). However, it appears as if Cole Kmet (Notre Dame) and Kylen Granson (SMU) have also been snubbed. In summary, going by receiving statistics alone, the top 8 Mackey Award finalists should be:

newtop8

I’m sure the committee is looking at more than just receiving stats, but I honestly have no idea what their criteria are. It doesn’t appear to be level of competition, as Giovanni Ricci of Western Michigan is on the list. However they are deciding it, receiving statistics should be a significant consideration. All in all, however, I don’t see how they justify leaving Pitts off the list.

An 8-Game Comparison of Feleipe Franks and Kyle Trask.

An interesting and albeit natural debate among Gator fans has risen regarding the QB situation for Florida. Kyle Trask gets a (deserving) amount of praise for his performance this year stepping in for an injured Feleipe Franks, who is (undeservedly) widely scorned. It seems to me that they are being graded in the court of public opinion on different curves. Don’t get me wrong- Kyle has been great and Franks certainly had some of those head-scratching moments. However, like most things sports, rationality isn’t the soup du jour.

As such, I decided to take a look at both QB’s performances across 8 games (since that is how many Trask has started now). I didn’t include the Kentucky game that was split between the two and just looked at the last 8 starts for Franks and did a side-by-side comparison. I looked at their performances along a few dimensions: Passer rating, completion percentage, yards, touchdowns, interceptions, longest pass plays. The results were interesting. The point here is not to determine who is ‘better’, just to look at the most recent comparable sample size and see if there is any difference in consistency as a passer (in terms of statistical variance). To determine who is better would require a much deeper dive, which I am not interested in- I’m glad they’re both Gators.

The Stat Lines

The table below shows each player’s overall performance for the 8-game sample:

traskstatline

franksstatline

Passer Rating

This is a metric taken from ESPN and isn’t the overall QB rating. I’m not sure how ESPN calculates this, but since I’m only analyzing these two as passers, I decided to include this statistic.

rating

This box and whisker plot shows that Trask is much lower variance than Franks. If you don’t already know, a box and whisker plot shows quite a few things. The top edge and bottom edge of the box are the hinges. These represent the 1st quartile and 3rd quartile. The area between them is the interquartile range (IQR). The line through the middle of the box is the median. The ‘X’ in roughly the middle is the arithmetic means (average). The whiskers represent the range of the data up to what is sometimes referred to as the upper and lower fences. These are the horizontal lines at the end of the whiskers. Any score outside of these fences is considered an outlier. From the graph above, we can see Trask has an outlier score (passer rating of 202 vs Towson).

Rating2

This graph shows the same data but has some different information. This shows how much difference it can make to include outliers in analyses. The arithmetic mean, or average, can be highly susceptible to extreme scores, especially in smaller sample sizes. When comparing Trask and Franks, it may not be best practice to do so by average passer rating because of the impact the outlier has on Trask’s score. The median may be a better statistic to use here, and they are virtually dead-even there (150.65 for Trask, 151.55 for Franks). Nonetheless, including Trask’s outlier, he is still much lower variance than Franks.

Yards

Yards

Yards2

Completion Percentage

comp perc

Outliers: Trask vs Towson (90%), Franks vs UT Martin (92.6%)

CompPerc2

Touchdown Passes

TD

TD2

Interceptions

Interceptions

The outlier is Franks’ 2-interception game vs Miami.

Int2

Longest Pass of Game

Long

Long2

My two cents

Trask is playing better. Not because he is throwing for more yards and more touchdowns, which is boosted by the fact that he is throwing more (32 attempts per game to Franks’ 24), but he is throwing more passes and is lower variance in every category except interceptions. He is the more consistent passer, and he is consistently good. Caveat- Running stats are not included and should certainly be considered, as Franks had 8 touchdown runs over that span and a median long run of 14.5 yards per game. Yes, he had a better line for 6 out of those 8 games, but it is what it is. He was a threat in the run game. I think both of them are good QBs, but the degree of separation is ever so slight. Performance-wise, I appreciate consistency and Trask has definitely been that while still being highly effective.

Extra:

Here is how FF 1st 8 games under Mullen would look when plotted against the above sample:

Ff vs franks

Ff vs franks2

Extra 2:

FF first 8 games under Mullen vs Trask:

FF1 vs trask2

ff1 vs trask

Charting success rate by downs and quarters: UF vs UGA edition

I’ve been charting Florida’s play-by-play success all season. I’m using this data for an end-of-season analysis but decided to throw the UGA game on here as a preview.

These quick charts are visual indicators of how well Florida performed by down and by quarter in terms of run and pass for both offense and defense.

Success is defined as follows:

1st down- Achieve 40% of the needed yards to convert or score.

2nd down- Achieve 60% of the needs yards to convert or score.

3rd & 4th down- Achieve 100% of the yards needed to convert or score.

*Kneel-down plays are omitted. Penalties, sacks, and turnovers are not counted as either pass or run plays (these are all being charted as ‘fail’ plays in general, and not presented here, but will be presented in the final analysis at the end of the season). The point here is to chart how often an executed play achieved its objective.

If you spot any errors, please let me know. I use a code to scrape the web and analyze the data. I QA the data as best as I can, but I would like to know if there are any errors so I can fix my code.

Florida Offense

uga.uf.off.down

Rushing Success Table by Down-Florida

Down # Rush Att. Success Fail Run success rate
1 6 2 4 33%
2 6 2 4 33%
3 4 1 3 25%
4 0 0 0 N/A

 

Passing Success Table by Down-Florida

Down # Pass Att. Success Fail Pass success rate
1 18 12 6 67%
2 8 3 5 38%
3 7 2 5 29%
4 3 2 1 67%

uga.uf.off.qtr

Rushing Success Table by Quarter-Florida

Quarter # Rush Att. Success Fail Run success rate
1 4 1 3 25%
2 3 1 2 33%
3 2 0 2 0%
4 7 3 4 43%

Passing Success Table by Quarter-Florida

Quarter # Pass Att. Success Fail Pass success rate
1 6 3 3 50%
2 5 2 3 40%
3 10 7 3 70%
4 15 7 8 47%

 

Georgia Offense

uga.uf.def.down

 

Rushing Success Table by Down-Georgia

Down # Rush Att. Success Fail Run Success Rate
1 16 6 10 38%
2 16 6 10 38%
3 5 3 2 60%
4 0 0 0 N/A

 

Passing Success Table by Down-Georgia

Down # Pass Att. Success Fail Pass Success Rate
1 11 8 3 73%
2 7 3 4 43%
3 13 9 4 69%
4 0 0 0 N/A

uga.uf.def.qtr

Rushing Success Table by Quarter-Georgia

Quarter Run Success Fail Run success rate
1 9 2 7 22%
2 11 6 5 55%
3 7 2 5 29%
4 10 5 5 50%

Passing Success Table by Quarter-Georgia

Quarter # Pass Att. Success Fail Pass success rate
1 9 5 4 56%
2 11 7 4 64%
3 8 5 3 63%
4 3 3 0 100%

 

Roster Talent and Win Percentage: The Non-Linear Relationship Between Recruiting Success and On-Field Performance in College Football

The recruitment of highly rated high school football players by college programs is big business. Major college football programs expend considerable budget in their efforts to recruit the most talented high school players. Blue-chip players, those rated by scouting services as 4 or 5-star players are relatively rare. As of October 10th, 2019, there were 347 Blue-Chip prospects according to the Composite 247 Ratings, the established industry standard for prospect scouting. There are approximately 1,006,013 high school football players in the United States (https://www.statista.com/statistics/267955/participation-in-us-high-school-football/

Some basic math lets us know that Blue-Chip players make up far less than 1% of all players. While validation for the rating methodology appears to be scant, the fact that so many big-time programs are recruiting these players offers logical support for the conclusion that these are indeed the better prospects. Coaches obviously want the best players, as that makes winning easier. The goal for this paper is to explore the degree to which recruiting success plays into winning on the field.

Among college football fans, it generally presumed that there is a direct (i.e. linear) relationship between recruiting Blue-Chip players and winning. In popular college football blogs and multiple articles, there are countless “analyses” of how “it is all about the Jimmy’s and Joe’s and not X’s and O’s”. Many die-hard fans appear to judge their teams’ recruiting rankings as a sure-fire indicator of whether the team is good or bad or even if the coach should be fired. The reality is that recruiting does not have a linear relationship to winning. But it does have a significant impact on winning- usually. This analysis found that in 3 out of 5 Power 5 (the most powerful) conferences in college football the on-roster talent was not statistically significantly correlated with winning percentage.

Overview

To conduct this study, roster talent ratings were obtained from the Composite 247 website for the top 50 most talented teams from 2016 to 2019 (as of October 25th, 2019). Each of the teams’ corresponding win percentage was then recorded. Each team was codified according to year and talent ranking. For example, the highest-rated team from 2016 was assigned a code of 2016_1. The 30th most talented team in 2018 had a code of 2018_30. And so on. Descriptive statistics and exploratory data analysis showed some interesting things at the conference level.

descriptives

What the above table shows us is the number of teams that were represented and their mean (average) roster talent score. So, over the last four years, the ACC, with 14 teams, has a possible number of 56 representatives in the data set (top 50 most talented teams over 4 years). They are represented 41 times (73%). The Big 12, with 10 teams (what does the ’12’ stand for again?) has 26 out of 40 possible representatives (65%). The Big 10 (14 representatives) has 39 out of 56 (70%), the PAC 12 (which actually has 12 teams- crazy) has 35 out of 48 (73%). The SEC has 55 out of 56 (98%). So the SEC has more, quantity-wise, talented teams.

The mean roster talent ratings are next. The Big 12 was the least talented group at 86.2, but virtually in a 3-way tie with the ACC and Big 10. Again, the SEC leads the way. Knowing there are some very talented teams in the other conferences (non-SEC), I wanted to look at how that talent is distributed among the conferences.

conf.histo

The above graph is a histogram for each conference. Not surprisingly, all of the other conferences are multi-modal whereas the SEC is approximately normally distributed. What this means is that the other conferences have more than one peak and valley, whereas the SEC is generally more bell-shaped (though quite a bit flatter here; low kurtosis but decent skewness). While there are certainly the ‘haves’ (Alabama, Georgia, LSU) and ‘have nots’ (Vanderbilt, Arkansas) in the SEC, the separation between the ‘haves’ and ‘have nots’ appear to be greater in the other conferences.

Regression Analyses

A regression analysis was then conducted for the entire sample. A simple linear regression found that There was a statistically significant correlation between roster talent and win percentage (p < 0.001, R2= 0.178).

overview

While the statistical significance indicates the correlation between roster talent and winning is unlikely to be a result of chance, the low R-squared value (a measure of effect size) indicates roster talent has a weak impact on winning percentage. It also indicates that the relationship may not be exactly linear. This was suspected to be the case, as the impact of roster talent may not have the same effect for each conference. Therefore, a polynomial regression analysis of the same data was conducted and found to be an improvement of the linear regression (p < 0.001, R2= 0.206).

overviewpolynomial

The polynomial regression was an improvement over the linear regression; however, the findings still indicate that roster talent has a weak correlation to winning on the field. At this point, it was necessary to look at the conferences individually to explore the relationship more granularly.

Conference by Conference

The below scatterplots show the same graph as above but filtered to only one conference at a time. The statistical output (significance and effect size) for each is below the graph:

acc

(ACC: p =0.08, R2= 0.164)

big12

(BIG 12: p =0.308, R2= 0.147)

big10

(BIG 10: p =0.017, R2= 0.247)

pac12

(PAC 12: p =0.386, R2= 0.092)

sec

(SEC: p <0.001, R2= 0.508)

Out of the 5 conferences, roster talent was only statistically significantly (at a standard threshold of 0.05) in the Big 10 (0.017) and the SEC (<0.001). The effect size for the Big 10 is 0.247, or 24.7%. For the SEC, it is .508, or 50.8%. This means that on-roster talent accounts for 24.7% of the wins in the Big 10 and 50.8% in the SEC from 2016 to present (October 25th, 2019). The key takeaway is that SEC fans are justified in fretting over recruiting success much more so than the other conferences.

A Deeper Look at the Southeastern Conference

The findings led me to want to look deeper into what is going on in the SEC. The plot below is the same as the SEC plot above but drilled down to the team level:

SECbyteamoutput

What this chart shows is that SEC teams generally perform to the relative level of their talent except for Tennessee and Florida. Tennessee generally underperforms while Florida generally overperforms except for 2017 when the Gators went 4-7. Kentucky’s 2018 year was a high achievement. An analysis was performed to explore how each team performed relative to their roster talent level for each of the years.

graph1

graph2

When summing the overall achievement for each team over the study period, Florida is the most over-achieving team in the sample.

achievementscale

If there is any doubt about Florida Head Coach Dan Mullen’s ability to maximize the talent on hand, it should be erased. The Gators are winning the most relative to roster talent even though they were 4-7 in 2017. Furthermore, Mississippi State is 3rd. A quick look at the SEC scatterplot shows us that MSU was overperforming when Mullen was a coach there but has since dropped off. Florida has done the opposite.

There is no doubt that the best place to be in is where Alabama is: maxed out on Talent and Wins, so there is no chance at really overachieving. After all, winning is what it is all about. And, in the SEC at least, recruiting success plays a big role in that. If you can’t recruit at an elite level, you’d better have a coach like Dan Mullen to give you a chance.

Additional:

As requested by Reddit user stevejust, I have added the team name and year to each of the conference scatterplots:

acc2

ACC

big10.2

Big 10

big12.2

Big12

pac12.2

PAC 12

*Added disclaimer:

The purpose of this study was to examine the linearity or lack thereof, between talent rating and winning percentage. The purpose is not to find the best model possible to predict winning- that is a worthy and separate endeavor! The primary curiosity was to explore if college football fans overstate/understate/state just perfectly their concerns about the recruiting performance of their favorite teams. The interesting find concerned the general lack of linearity between quantified talent level and winning. I have received considerable feedback (virtually all of it positive, which is awesome) and several suggestions regarding other potential variables to consider, models and methods that may be superior or an improvement over the polynomial regression. Each of these that I have read has been, to varying degrees, perfectly legitimate and well-thought-out. However, it is important to emphasize, that this study was not about the model of best fit- it was about the supposed linearity between talent ratings and winning. Thank you to all who have commented and provided excellent ideas for future approaches. I am learning from you and gathering great ideas. Cheers!