Size as a Predictor Matters

[OldNavy]

The Alignment for the 2009-2010 and 2010-2011 seasons put 184 teams in 3A. The average size school of those 184 teams was 676.7 students. There were 82 teams larger than average and 102 teams smaller than average. 52 of the larger than average schools made the playoffs or 63.4% of larger schools qualified. 44 schools in the smaller than average category qualified for the playoffs for 43.1% of smaller schools going to the post season.

The average size of D1 schools making the playoffs this year is 840.66 students. The smallest 16 schools in D1 averaged a student population of 756.8. For comparison the largest 32 schools in D2 averaged 745.3 students. The smallest 32 schools in D2 averaged 524 students.

By these averages it would seem to make more sense to send the largest two teams in a district to D1 and the smallest team from a district to D2. The largest schools with greater numbers and depth should be able to play the six playoff games better than the smaller schools.

Last year’ champions, D1 Henderson, with 879.5 and D2 Carthage, with 731, were both larger than the average school in 3A. The average size for a 3A champion was 805.25.

I previously posted the Massey predictions for the teams in the playoffs. It successfully picked 77% of the winners. Then I looked at the school size. Admittedly, some of these margins are narrow, however if you chose the larger school to win over the smaller school you would have been right 58.3% of the time.

The school size and their first round results are listed below. Does 58.3% of larger schools and 41.67% of the smaller school winning over the larger school sound like a familiar number?

Division I
School Student Population (Score)
Larger school wins (in red)

REGION I BI-DISTRICT
Cooper 808 over Borger 771 (34--9)
Seminole 626 over Fabens 726.5 (61--0)
Abilene Wylie 940 over Burkburnett 869 (28—17)
Alvarado 974 over Mineral Wells 961 (69—6)


REGION II BI-DISTRICT
Celina 558 over Sanger 735 (42—21)
Lovejoy 946 over Dallas Madison 773 (24—7)
Lindale 987 over Kaufman 950 (38—28)
Henderson 879.5 over Paris 890 (55—17)

REGION III BI-DISTRICT
Chapel Hill 870 over Jasper 770 (70—28)
Levaga 772 over Palistine 825 (35—20)
Huffman Hargrave 924 over Silsbee 853 (19—14)
Navasota 801 over Sealy 756 (40—13)

REGION IV BI-DISTRICT
Lampasas 987 over Boerne 887 (35—21)
La Vernia 892 over Somerset 932--------------(28—0)
Columbia 875 over Rockport Fulton 981-------(34--15)
Raymondville 600 over RGC Grulla 950 (9—29)

DIVISION II
REGION I BI-DISTRICT
Shallowater 433 over Dalhart 489 (40—7)
Midland Greenwood 517 over Pecos 609 (42—14)
Snyder 714 over Graham 732 (35--19)
Kennedale 924 over Decatur 857 (24—14)
Lubbock Estacado 799.5 over Perryton 584 (61—21)
Monahans 565 over Clint 588 (43—6)
Brownwood 877 over Bowie 432 (49--13)
Glen Rose 518 over Bridgeport 619 (57—56)

REGION II BI-DISTRICT
Argyle 611 over Commerce 445 (49—7)
Prosper 870 Wilmer-Hutchins 715 (28--7)
North Forney 904 over Wills Point 820 (34—28)
Carthage 731 over Liberty Eylau 719 (28--7)
Van Alstyne 439 over Whitesboro 459 (40—14)
Princeton 797 over Roosevelt (http://topics.dallasnews.com/topic/Roosevelt%2C_NY) 758 (30—21)
Van 682 over Ferris 640 (51—23)
Gilmer 628 over Atlanta (http://topics.dallasnews.com/topic/Atlanta%2C_GA) 509 (48—14)

REGION III BI-DISTRICT
Bullard 594 over Center 617 (24-14)
Waco Robinson 684 over Mexia 541 (28—7)
West Orange-Stark 699 over Cleveland 886 --------(58-14)
Columbus 535 over Smithville 527 (21—14)
Rusk 560 over Kirbyville 432 (47-18)
Lorena 481 over Fairfield 542 (28—14)
Coldspring-Oakhurst 503 over Bridge City 714 (7—0)
Bellville 673 over Giddings624 (28—18)

REGION IV BI-DISTRICT
Liberty Hill 746 over Navarro 461 (35—14)
Devine 538 over SA Houston 795 (27—19)
Wharton 583 over Ingleside 558 (36—35)
Orange Grove 504 over Port Isabel 651 (28—14)
Wimberley 681 over Taylor 882 (38—31) 2OT
Gonzales 657 over Hondo 632 (21—10)
Sweeny 587 over Sinton 565 (35--7)
Rio Hondo 641 over Lyford 446 (27—14)

[Leopard4Life]

The bigger school won more than 50% of the time and the smaller less than 50%, but it was within a plus/minus 10% margin. Flip a coin a hundred times and it probably won't even come this close. Thanks for doing all this research and gathering this data, however I don't think the results showed that size matters. I've always wondered about the size of the HS program, like 150 kids vs. 75 kids. I think larger programs are more successful. Lovejoy has about 145 in their program and Celina has about 150, where Farmersville has about 75-80. It is harder for smaller programs to compete.

[DavidWooderson]

Exactly........Larger schools have more to chose from. A lot of times can have different players for offense and defense or majority of the players. That helps.

[LionFan72]

Size matters,............where have I heard that before???????????????

[RPF2666]

Size matters,............where have I heard that before???????????????

Keep your 'private stuff'...private...:)

[RPF2666]

The bigger school won more than 50% of the time and the smaller less than 50%, but it was within a plus/minus 10% margin. Flip a coin a hundred times and it probably won't even come this close. Thanks for doing all this research and gathering this data, however I don't think the results showed that size matters. I've always wondered about the size of the HS program, like 150 kids vs. 75 kids. I think larger programs are more successful. Lovejoy has about 145 in their program and Celina has about 150, where Farmersville has about 75-80. It is harder for smaller programs to compete.

It is completely logical and legitimate, that if YOU have 500 boys and 150 play football, and if I have 250 boys and 150 play football, that YOU are likely to have
fewer who play because 'everyone else does' and more who play because football really is their game. Size will ultimately matter over the long run, or they wouldn't split the schools into SIZE classes, they'd just do it strictly by logistics.

[OldNavy]

The bigger school won more than 50% of the time and the smaller less than 50%, but it was within a plus/minus 10% margin. Flip a coin a hundred times and it probably won't even come this close. Thanks for doing all this research and gathering this data, however I don't think the results showed that size matters...

I would agree with you except there are 102 small schools and 82 large schools, so the percentages are already skewed. The 60% to 40% win ratio that I saw last year is holding true again this year. School size obviously is not the only factor, but it is a large (pardon the pun) factor.

[LionFan72]

I would agree with you except there are 102 small schools and 82 large schools, so the percentages are already skewed. The 60% to 40% win ratio that I saw last year is holding true again this year. School size obviously is not the only factor, but it is a large (pardon the pun) factor.

Logically, I have to agree with you on merit, but care to explain the Celina dominance over last 10 years, or the rise to dominance by South Lake Carroll over their rise to the top thru all classes?? Sometime is just not the size of the dog, but the size of the fight in the dog and a great supporting Coaching staff!

[JustAFan]

Wow, this would be a great project for a student in Stats! Because you have to look at significance, not just numbers. Is the difference significant in Carthage 731 over Liberty Eylau 719 (28--7)? Or Columbus 535 over Smithville 527 (21—14)? Let's be realistic! Yes, there will be a statistical significance between Brownwood 877 over Bowie 432 (49--13) since Brownwood has more than twice the students to choose from. But, it makes it really difficult to explain Raymondville 600 over RGC Grulla 950 (9—29). Your stats may be correct, but they may not have any actual mathematical significance. It is kind of like the announcers during the World Series making a comment that "St. Louis leads the major leagues in hitting a single between the 1st and 2nd basemen when there are men on 2nd and 3rd during the 4th inning and they are behind by 3-4 runs in a series environment where they are losing 2 games to 3." No, that was not said, but that is one sport where they come up with some of the weirdest stats, that have no impact on reality. And this is very similar.

[Leopard4Life]

It is completely logical and legitimate, that if YOU have 500 boys and 150 play football, and if I have 250 boys and 150 play football, that YOU are likely to have
fewer who play because 'everyone else does' and more who play because football really is their game. Size will ultimately matter over the long run, or they wouldn't split the schools into SIZE classes, they'd just do it strictly by logistics.

There are probably a lot of schools that have the 'everyone else does' issue. However in larger schools football program loses good players/athletes to other sports. The more options/sports a school has the less likely that all the quality athletes will be in the football program. There are more 1 sport athletes than ever before. If a large school got all there top athletes and took the top 150 compared to a small school doing the same thing then your point is 100% valid. If the top athletes came out of basketball, baseball, wrestling, soccer, lacrosse, cross-country, track, tennis, swimming, etc... then the quality of the football program would increase, but those other sports would decline.

I agree that size could be a factor, however these numbers are too close to tell. If it is a 60/40 split between large/small school and these numbers came out 58/42 winning/losing then it is to hard to tell because both are close to plus/minus 10%. If 80% of the schools were large compared to 20% being small and the large school won 80% of the time it would validate the theory. Since all the numbers are near 50% it is incunclusive.

[BILLYFRED0000]

I am not going to belabor this point but size does matter. In Celina's case we have a middlin to small school with outstanding participation and a stable program going back 23 years. We are the exception. Look closely at D2 and see how many larger than the average size are competing and playing. d1 does not matter so much because many d2 districts have several large schools.
Argyle 611 over Commerce 445 (49—7)
Prosper 870 Wilmer-Hutchins 715 (28--7)
North Forney 904 over Wills Point 820 (34—28)
Carthage 731 over Liberty Eylau 719 (28--7)
Van Alstyne 439 over Whitesboro 459 (40—14)
Princeton 797 over Roosevelt 758 (30—21)
Van 682 over Ferris 640 (51—23)
Gilmer 628 over Atlanta 509 (48—14)
In this group of schools there are only 6 truly small schools as I start that at about 630 and down.
And that is not arbitrary when you consider that small schools should at least by definition be below average when they comprise over 60 percent of the schools. You could take Princeton for example. 797 vs VA at 439. almost twice the size. Inside one region of DII. Size does matter. A large school with athletes has a good chance against a small school with a team. Depth is hard to beat.

[waterboy]

In 2009, the D-1 champion (Gilmer) had 628 students, while the D-2 Champion (Carthage) had 731 students. How weird is that?

[BILLYFRED0000]

exactly...... it all depends on how the larger schools shake out in their district instead of what schools are larger.....

[OldNavy]

Wow, this would be a great project for a student in Stats! Because you have to look at significance, not just numbers. Is the difference significant in Carthage 731 over Liberty Eylau 719 (28--7)? Or Columbus 535 over Smithville 527 (21—14)? Let's be realistic! Yes, there will be a statistical significance between Brownwood 877 over Bowie 432 (49--13) since Brownwood has more than twice the students to choose from. But, it makes it really difficult to explain Raymondville 600 over RGC Grulla 950 (9—29). Your stats may be correct, but they may not have any actual mathematical significance. It is kind of like the announcers during the World Series making a comment that "St. Louis leads the major leagues in hitting a single between the 1st and 2nd basemen when there are men on 2nd and 3rd during the 4th inning and they are behind by 3-4 runs in a series environment where they are losing 2 games to 3." No, that was not said, but that is one sport where they come up with some of the weirdest stats, that have no impact on reality. And this is very similar.

I agree with you that for one game when one school has 640 and another has 650, size probably doesn't matter. However, if you look at the universe, and for the last two years, 60% of the larger schools go to the play offs and only 40% of the smaller schools go and in the playoff games without any other consideration 58% of the larger teams win and only 42% of the smaller. At what point does it become significant?

I am not trying to explain why one smaller team wins over a larger team. It is similar to fliping a coin. If you flip it once, and it comes up heads. Then heads is all you got. But if you flip it 1,000 times you get an average of 50% heads and 50% tails. We had about 920 games in district play and more than 60% of the larger teams made the playoffs for two years in a row. I have not looked at a large number of seasons but there is enough data here to make me believe it is not a statistical anomaly.

[OldNavy]

OK, 52 of the 82 larger than average schools made the playoffs. Only 44 of the 102 smaller than average schools made the playoffs. The larger than average school won over the smaller schools nearly 60% of the time and both state champions were larger than average and both won over smaller schools. If my memory serves, these are almost the same results as last year.

Some have argued that this is not statistically significant. I think it is significant and I think the UIL does also. I think that is why they are pushing the big school, small school districts and playoffs. In the most recent alignment, 87 schools are larger than the average and 103 are smaller. Many of the dropdowns from 4A are pretty potent and some of those will go D2 because they are not the largest in their district.

[ziggy29]

It may not seem *extremely* significant in terms of what happens in the playoffs, but looking at the OP, it looks like the size difference tends to be small (usually less than 100) in the playoff match ups, with a handful of exceptions. But I think it might be more interesting to see how it impacts the chances of getting to the playoffs at all. When 3A ranges from 450 to over 1000, it doesn't seem like those closer to 450 have much of a chance.

If a team that barely makes the minimum in 3A is up against a bunch of schools with enrollments over 800-900 in district, what chance does it have?

Yes, I'm bitter and yes, I'm whining. I admit it.

[BEAST]

It may not seem *extremely* significant in terms of what happens in the playoffs, but looking at the OP, it looks like the size difference tends to be small (usually less than 100) in the playoff match ups, with a handful of exceptions. But I think it might be more interesting to see how it impacts the chances of getting to the playoffs at all. When 3A ranges from 450 to over 1000, it doesn't seem like those closer to 450 have much of a chance.

If a team that barely makes the minimum in 3A is up against a bunch of schools with enrollments over 800-900 in district, what chance does it have?

Yes, I'm bitter and yes, I'm whining. I admit it.

Celina has done quite well with low numbers.




BEAST

[Saggy Aggie]

It may not seem *extremely* significant in terms of what happens in the playoffs, but looking at the OP, it looks like the size difference tends to be small (usually less than 100) in the playoff match ups, with a handful of exceptions. But I think it might be more interesting to see how it impacts the chances of getting to the playoffs at all. When 3A ranges from 450 to over 1000, it doesn't seem like those closer to 450 have much of a chance.

If a team that barely makes the minimum in 3A is up against a bunch of schools with enrollments over 800-900 in district, what chance does it have?

Yes, I'm bitter and yes, I'm whining. I admit it.Lol well just to be completely honest, I don't think llano would be competitive if they had 1200 kids

[OldNavy]

Celina has done quite well with low numbers.


BEAST

Yes they have, but the numbers don't keep a small school from doing well, it just places the odds a bit less in their favor. I also think a small school doing very well is exceptional, and maybe more than a little bit lucky. There are a lot of things that could be studied. Do larger schools afford and hire better and even more coaches? Can they afford better facilities for practice and game preparations. Obviously there is greater probability of having genetically gifted students in the larger populations. There may be an intemidation factors too. Some schools have bands half the size of some of the smaller schools in 3A and can put twice as many players on varsity. That has got to have some impact.

It does not keep the smaller school's 11 players from beating the larger 11, but over a season, as a general rule, injuries will take their toll on the smaller school. I will use Chapel Hill as an example. They played many games against good competition without as many as 8 starters. Their depth was just greater than most schools, much less those with half the population.

[Sville]

Personally I think they should have adjusted the numbers to have about 200 teams in 3a and take 4 from each district, two D1 and two D2. Just like 4a and 5a. I think this would of helped quite a bit.

[waterboy]

Personally I think they should have adjusted the numbers to have about 200 teams in 3a and take 4 from each district, two D1 and two D2. Just like 4a and 5a. I think this would of helped quite a bit.
I can agree with this. Why is it that 3A only has 3 teams that make the playoffs when the rest of the classifications have four? If they did change the numbers just a little and do a split of some sort it would make it more fair. 3A football is the best and purest classification in the state, in my opinion, yet if a team goes D1 they don't get the same exposure in the state championship game because it's played a week earlier than the rest. The logistics of the changes would probably be a nightmare, though, and that's the only reason I see for such a split to not come to fruition. If they can't split 3A because of the logistics they should add another team to the playoffs.

[CHS_89]

Nice stats...thanks Old Navy....In some other states, the districts are decided for male sports based on how many male athletes are in the school. Just wondering how that would work in Texas? Would probably be a nightmare.