Florida Gators Blog

Roster Dispersion of Major College Football Teams: From Where Do They Get Their Players?

This study will be an ongoing one, as this is just preliminary. The below displayed maps show where some major college football teams accumulate their players.

18Gators

Florida Gators

18canes

Miami Hurricanes

18nulls

Florida State Seminoles

18buckeyes

Ohio State Buckeyes

18tide

Alabama Crimson Tide

18dwags

Georgia Bulldogs

These maps show the entire roster makeup, so it includes walk-ons, etc. More work is to be done here, but it is at least kinda neat (to me) to see from where teams obtain their players.

WLOCP hangover?Florida, yes. UGA, no.

From Reddit user agage3

“Our performance after a WLOCP game win compared to our performance after a WLOCP loss. I swear when we lose our season just tanks but when we win that game we ride the wave to consecutive late season wins.”

To see if UF truly has a hangover after a loss at the WLOCP, I reviewed the results of the game since from 1998-2017. I looked at how Florida and Georgia each did the following week as well as for the rest of the season. Florida has gone one to win 64% of their games overall after the WLOCP in that time frame with Georgia winning 75%.  While, the game itself tends to go our way, we fair worse afterwards.

fl vs ga

For the statistically-interested, an independent samples t-test was conducted to determine if the mean winning percentage between the two was statistically significant. The results are statistically significant at the a= .10 (p= 0.82). Levene’s test was not significant, thus equal variances were assumed. Effect size was small to medium (r²=0.08).

Takeaway, though not strongly significant (typical a level is 0.05, and medium effect size for r² is 0.09), the dwags have had slightly more success after the WLOCP over the last 20 years, but Florida has the advantage where it counts: head to head and championships.

Another Look at Gator QBs since Tebow- a statistical breakdown

Every single Florida Gators football fan has heard it a million times… quarterback woes since Tim Tebow left after the 2009 season. As requested, I took a look at what has been going on with Florida’s QBs after Timmy Terrific moved on from Gainesville.

I used the key metrics for quarterbacks- completion percentage, yards, touchdowns, and interceptions. I don’t use the rating because it is comprised of these same stats and would be redundant. Plus, I’m not a fan of how ratings are computed, but that is for another discussion. I then took Florida’s quarterbacks statistics (passing only) that have played since Tebow left. I projected each of these players’ stats over a 12-game range to give an idea of what a full season with each of them would have looked like. Here is what we get:

qb chart 1

I then did the same for Tim Tebow for the 3 season he started at Florida. Because he played in more than 12 games in those years, his projected numbers are actually lower than the actual stats he put up. But, hey, we have to put them all under the same light to see how it shakes out.

qb chart 2

Since these are passing numbers only, they do not take into consideration the impact Tebow had as a rusher, which was significant. No need to pile on… he was a hell of a passer as well. I want to point out that not a single quarterback since Tebow would have projected to put up any statistics matching his in any category except Luke Del Rio, who projected to have 2,716 passing yards over a full season in 2016. Treon Harris would have thrown 5.3 projected interceptions in 2014, a little lower than Tim’s 5.5 in 2007. Yes, I know it is because Treon likely would not have thrown very many passes, but just pointing that out.

Ok, so we know Tebow was good and he probably left us all spoiled. Perhaps this throws our expectations off. Well, to figure this out I took all the current power 5 quarterback passing statistics (same ones used above) and used those as a standard by which to compare our Florida samples. Then each of the quarterback’s projected scores were standardized and compared to today’s current scores.

qb chart 3

When you standardize scores, you are centering them on an average (mean) of zero. When you look at the above chart, each score then moves negatively or positively away from zero in uniformed increments (standard deviations). This is a common statistical method for taking values along different dimensions (like yards vs TDs) and putting them on equal footing for analysis. Also, more than 1.5 standard deviations in either direction gets into outlier status, and 3 standard deviations from the mean are considered extreme… out of the 99th percentile. Anyways… in the chart above negative numbers indicate how far below the average the player was in that category (good for interceptions, bad for everything else). Positive numbers of course are the opposite.  And to reiterate… those findings are how each QB, projected over a 12-game season, would compare to the current power 5 quarterbacks’ statistics. The 2009 Tim Tebow would be throwing for 1.08 standard deviations (which is currently equal to about 561) yards above national average for power 5 QBs. He would also have 4.3 more touchdown passes than the average current QB, which would equate to about 18. I digress…

There is an excellent argument that could be made that Tebow benefited from being surrounded by ridiculous talent. The argument would be good. So, let’s take a look. The team talent column below is the 4-year moving average for the Florida Gators recruit talent ratings as taken from 247 Composite.

qb chart 4

As we can see, Tebow’s teams were good (obviously), but not necessarily out of the range of some of the other teams. To look at it from another direction, I looked up each team’s strength of schedule (SOS) ranking for that year (taken from teamrankings.com). There are several different SOS rankings to go by, and they probably vary significantly. However, under this one metric as a constant, it looks like things have gotten easier for Florida, schedule-wise, since Tebow left. In all likelihood, this is probably because winning teams get to play more games against other winning teams, driving up the difficulty level.

Another question to ponder, is were any of the quarterbacks since Tebow expected to be any good? Were they highly regarded, blue-chip recruits? Yep. They were.

qb chart 5

Looks like Florida has had some pretty good blue-chip players at the quarterback position. Ironically, it was the non 5-star players who have generally done pretty well. Grier is obviously an excellent player (4 stars) and performed above average for Florida across the board. Del Rio was on pace to have a good year yardage and touchdown-wise, but a bad, bad year in terms of interception. Other than that, Florida’s quarterbacks have largely played below average.

In closing, having a quarterback like Tebow probably did skew our expectations for quarterback performance. However, to compound Florida fan misery, the quarterbacks since Tebow have not done well compared to other schools’ passers. This makes the contrast even more marked. Go Gators.

Florida’s Penalties, We love ’em (at least on Offense).

Question from Reddit user tacobellemel

“If Florida is penalized at a statistically significant rate higher than other schools. I found data back to 2002 and we are always in the bottom third in fewest penalties. Interestingly enough I also noticed FSU was usually down there with us so I wonder if there is some sort of unconscious bias against Florida schools.”
First the data. I collected data from 2000 to 2018 for the average number of penalties on both the offensive and defensive side of the ball for all FBS teams.
When compared to the average, Florida consistently gets more penalties on offense the the average team in college football.
Defensively, Florida is a mixed bag and average out to the rest of the teams.

UF Penalties

To throw some stats in, here is a bell curve with Florida’s offensive penalties, where florida is worst than about 99.3% of teams in the FBS.

Bell Curve Penalties

Now for some good news, most stat heads agree that a negative yardage play on first down is significantly detrimental to that drive’s success. If I remember correctly, a team that takes a 5 yard penalty on 1 & 10 punts 25% more than teams that avoid 1st down errors.

Florida is significantly better than most teams on avoiding penalties on first down. Below is the bell curve for 1st down penalties,where Florida is about 71% better than most teams on first downs.

UF Penalties 1st Down

Go Gators (except on 2nd & 3rd down)

—–Matthew Mann

 

Third and Grantham: An analysis of Todd Grantham’s history on third down

I took a look at Florida Gators’ defensive coordinator Todd Grantham’s historical performance on 3rd down. I also looked at the talent level he had for each of the years he was a defensive coordinator. An important note is that 3rd down conversion rate (opponent) is obviously not an indicator as to how well the defense plays overall, so this is not to be an analysis of Grantham’s worth as a defensive coordinator. The chart below shows the overall data.

grantham

What the chart is showing is as follows (and a helpful hint is that the negative numbers are better):

  • Prev. yr. conv. Rate is the standardized score of opponent’s conversion rate from the year before. So, for the 2005 Cleveland Browns, this is how well the Browns did in allowing 3rd down conversions in 2004, which was the year before Grantham was their defensive coordinator. The -0.386 score means they allowed slightly fewer than the league average (standard deviations from the mean) conversions.
  • The next column over, the score is .706. This means the Browns allowed more 3rd down conversions compared to the average amount for the league. The ∆ con rt (∆ is a mathematical symbol for ‘difference’ for all of you who haven’t had math in a while) indicates the rate that the team’s performance on third down changed from the previous year to the year under review (in this example, 2005).
  • The 1.091 means the Browns allowed over one full standard deviations more 3rd down conversions in 2005 than they did in 2004.
  • The value of that is reflected in the next column labeled ‘Improve?’. As most of you will be able to figure out by the comment “Got worse”, you will be able to see that Grantham’s defense fared much more poorly in defending 3rd down than the previous year.
  • Overall rank is pretty straightforward. The Browns were 27th in the NFL (out of 32 teams) in allowing 3rd down conversions. For college football, the overall rank is out of all D1 teams (130 or so). I could’ve cut this down to P5, but that was additional work, so I left it as is. As long as it is constant, the rank has the same overall value IMO.
  • The talent score is a normalized score the teams’ 4-year moving average scores from the 247-composite analysis. So, what that means is that in the first year Grantham was at Georgia, the Bulldogs’ roster was in the op 88th percentile for the nation in terms of recruit talent.

I will let you guys evaluate the value of the information… I think my feelings are made clear regarding the Gators this year by looking at the chart. In 2017 Florida was 10th in the nation allowing only 4.2 3rd down conversions per game (probably because our opponents were never in 3rd down…). This year we are currently allowing 6.6 conversions per game, good for 115th in the nation. The last 2 games have significantly impacted this, so keep that in mind. Also, UCF is 116th in the nation, allowing 6.9 per game.

Also, this study will continue to go further in-depth. We are attempting to gather the game-by-game blitz information and play-by-play results which will give us a more micro-view of what is going on. For now, this overview will give you an idea of how Grantham has historically done on 3rd down.

Champs vs Chumps: A comparison of how they do in their first four years at the helm.

This study proposes to examine how recent national championship coaches faired in their first four years with their team that would go on to win a championship versus coaches who finish 25th in the final AP poll. What are the conditions in which each of the two groups took over? How fast did each of the two groups achieve success and how well did they do in recruiting compared to each other?

For the data, I took the ten most recent national champions head coaches. And the ten most recent coaches of the AP poll that finished ranked 25th. The goal was to see if there was any discernable difference in how these two groups performed in their first four years. I chose four years because after that there appears to be a lot of turnover at the coaching position. A key concept to keep in mind is that the championship coaches did not necessarily win a natty in the first four years (for example, Mack Brown of Texas and Dabo Swinney of Clemson took 7 and 8 years to win their first championships, respectively). Another important note is that the study measured how the coaches did in their first four years with their championship team, and not their first four years as a head coach overall. As such, Nick Saban (LSU and Alabama) and Urban Meyer (Florida and Ohio State) were in the sample twice. An interesting note is that Auburn’s Gene Chizik was in both data sets, as he won a title in 2010 and finished 25th in 2011.table 1

The table above shows each of the last ten championship coaches, their first year with the championship team and the year they won their first national championship. The table below shows the coaches for the 25th group.

chumps

The ‘Champs’ and the ‘Chumps’, interestingly enough, both took over teams that were (on average) 6-6 the year before. However, the Champs went about winning more and steadily from there. The chart below shows this.

champs 1

From the chart above, you can see that both groups started at the same spot. Year with new team marker #1 is actually the year prior to their taking over. Both groups were at .500. From there the Champs (blue line) rose more quickly and kept going up. The chumps rose in winning percentage but dropped off considerably after year 3.

What about talent and recruiting? That has to factor in, right?

Yes. Using the four-year moving average for recruit ratings to determine the general level of measurable talent on each of the teams, I tracked the progression from starting point and onward. The Champs typically took over slightly more talented teams (average rating of .739 to .706) but did a better job of infusing talent continuously:

talent

Here, you can see that the champs started in a better position in year one and moved their recruiting averages up steeply before and never dipping to the level of the chumps. The scores are shown in the table below.

table
The colors in the table headers are not intended to represent Florida or fsu colors. Pure coincidence…

So, what does this mean we should expect from UF coach Dan Mullen? Only two Championship coaches took over teams that had 4 or fewer wins the previous year. Saban at LSU (3-8) and Brown at Texas (4-7). If you give Florida the extra win that would’ve happened had their game against whoever was cancelled and say they would’ve been 5-7 in Jim McElwain’s last year, then you could add Chizik to the list of coaches that took over 5-win teams (Auburn was 5-7 the year before). This indicates Mullen may have a little more of a project on his hands than the usual championship-winning coach. But Mullen does take over a team with decent talent and recruited well talent-wise his first year (Florida has the nation’s 14th most talented roster and talent rating average of 90.75 in his first recruiting class). I will let the reader draw the conclusions as to what is fair for Mullen to accomplish in his 1st year. But I will say this- according to recent history, the first four years show a significant upward trend in wins and recruiting for coaches that win championships versus those who faltered.

The Florida Gators vs the South Carolina Gamecocks: How they match up.

The Gators will host South Carolina this Saturday in Gainesville. To analyze how the two teams match up with each other, I took a look at the most relevant statistics on offense, defense, and special teams as they apply to a head-to-head analysis. The main components looked at were offensive scoring per play, defensive scoring per play, field goal percentage, team talent, and achievement level of the team relative to their talent, which is a reflection of coaching. I utilized a series of data transformations and standardization to account for the weight of each variable. This puts all of the variables on even footing and minimizes distorting the results.

sc vs uf

The above chart shows the breakdown. Florida has an adjusted advantage of 0.09 in scoring offense, 0.429 in scoring defense. The kickers are equal. Florida has a slight talent advantage. The talent is calculated by averaging out the 247 rating for all players with ratings on the rosters. Of note, this includes any player that has a talent rating, whether on scholarship or not. The ‘achieving’ category is obtained by comparing win percentage to the talent on the team relative to the rest of the SEC. This is done to measure how each team is doing within their group. Of course, some have harder schedules than others, but this serves as a good baseline. Both coaching staffs have their teams out-performing their talent (even though Florida didn’t do that last week). By the way, I am confident in this metric, because in the overall analysis of the SEC this year, Kentucky is the most overachieving team (wins to talent ratio). All of these factors give Florida a slight (z=0.264) advantage.

The real scary part, however, is the wildcard factor. For this game, it is the quarterbacks. Because QB play impact both offensive and defensive scoring, it isn’t included in the original model. Redundancy (known as multicollinearity to you other stats folks) can adversely affect the outcome. However, the difference in quarterback passing effectiveness (QB S5) is strongly in South Carolina’s favor. Jake Bentley has played about 41% better than Franks (this is derived from the empirical rule). Again, this is redundant- Bentley’s play has contributed to the offensive scoring per play for USC, just as Franks’ has for Florida. But this disparity is concerning to me. That being said, the Gators should win by a small margin at home. If Florida gets good to decent quarterback play, they should have no problem at all.