Public 106 Private 62

[Snetsrak61]

@ramblinman so what is your issue with #2 and #4 did I not use enough adjectives for you?

And please tell me a private school that has not benefited from the 30 mile radius?

Or do you need me to specify with...what successful private school that has not benefited from the 30 mile radius.

Well, you could counterpoint and say the 30 mile radius is a limitation put on private schools. Then you can question if the limitation is enough to encourage equitable opportunity in athletic competition.

But very technically speaking it is actually a limitation. A benefit? Eh.

But perhaps removing any radius, and inserting a multiplier of increasing degree based on the student body IS the answer. ¯\_(ツ)_/¯

 

[Alexander32]

For a "mathematician"/"statistician", you leapt to 100% certain rather quickly, especially since the the original hypothesis used such vagaries as on average and equal. You didn't really define what equal meant, to what precision the data would be measured, etc. It was a disingenuous test and just like all the other posters, you found data to match what you wanted to say, only you used your "credentials" to lend more authority to your conclusions.

Further, you might consider shortening your posts, as they tend to be repetitive statements within the same post, which means people stop reading.

* To be clear, the hypothesis being tested was whether or not the private high school playoff qualifying teams are equal in ability to the public high school playoff qualifying teams.

* By equal I mean the average probability of both sets of teams (public and private) winning a game between them is identical. By definition that means a 50/50 chance.

* The definition of average being used here is the statistical mean.

* By using the phrase "on average", I was trying to convey the concept that not every game between the two sets of teams would be a 50/50 proposition, but that the average of all the games would be 50/50.

* By stating a statistician would conclude with a 100.00% level of certainty, the inference is that the precision to which the data was measured was 1/100 of one percent.

* I didn't want to say anything. I allowed the data to make the statement.

* I did not develop the data. I did not search for the data. The data used was provided by someone else and was offered to everyone reading this particular thread.

*** Let us disregard the original data and use the data provided by Snetsrak61 instead. It seems most of us agree Snetsrak61 is an unbiased source.

*** Let us disregard the calculations of Alexander32 (me), which you seem to believe are biased. Let's use a calculator developed by Maciej Kowalski, a PhD candidate who is not involved in this discussion/debate. It can be found at https://www.omnicalculator.com/statistics/coin-flip-probabilty/

One may have to click on the box labeled "Statistics", and then a second box labeled "Coin Flip Probability Calculator" in order to reach the calculator.

If you were to enter 1,000 for the "Number of flips", change "exactly" to read "at least", and then enter "650" for the minimum number of "heads" desired, the calculator will determine the chance of that outcome occurring randomly. [650 is 65% of 1,000, and was the percentage of games won by the private schools according to Snetsrak61's data.] You will see the chance of this outcome occurring is 0%. If one enters 600 for the minimum number of heads desired, it can be seen that the chance is less than one chance in a billion outcomes. Given that the chance of the private schools winning 65% of the games is next to impossible, unless they were better (as a group) than the public schools, one can conclude with at least a 100.00% level of confidence that they were better.

This result will demonstrate, to an objective individual, that the set of private schools participating in the playoffs over the past two decades was better than the set of public schools participating in the playoffs over the past two decades. One might then conclude that the multiplier is justified.

It is true, as pointed out by Snetsrak61, that for both the private schools and the public schools the better teams are weighted more heavily by the data. This is true because the better teams will advance further in the playoffs, and therefore will have played more games. That fact is actually desirable. Most contributors to this thread are not overly concerned about teams winning a playoff game here and there, or even reaching the quarterfinals from time-to-time. A concern arises, if it arises at all, when teams start dominating a particular class level. Therefore, one would want an analysis of competitiveness to weight the better teams more heavily.

Although the previous paragraph seems reasonable, we still would not want a single team to skew the results too heavily. That will not happen. The best that any one team could do is to win a championship in all 21 years comprised in the analysis. In order to do that, the team would need to go 105 - 0 (5 victories x 21 playoffs). They could not account for more than 105 victories. In Snetsrak61's data, the private schools won 879 games. Therefore, even a team that won 21 straight championships would not account for more than 12% of the private-school victories. Besides, if that ever did happen (21 straight championships by a single school), I imagine even the private-school supporters would agree that does not represent a reasonably competitive situation.

 

[Alexander32]

So FWIW, here is what my initial evaluation of the playoff W% data is. I'm also doing 2001-2022, and I am not classifying open-boundary publics with the privates. I could check the numbers with that change next, but I think it'd only make the analysis further off than Antioch's

Overall Private v Public Record
879 - 471, for a 0.651 win percentage in advantage of the privates.

Each Class with Share of 50% win seasons for the publics (first column includes exact 0.500, second column excludes). And total win % for private schools. Especially on the small schools, this looks way off from what Antioch twitter posted.

1A - 16   14   0.423
2A - 8    6    0.617
3A - 10   9    0.570
4A - 5    4    0.631
5A - 2    1    0.793
6A - 5    2    0.709
7A - 6    2    0.676
8A - 5    2    0.684

One interesting aspect of this I think is still the "have" v "have-not" debate. There are about 180 public schools who have lost to, but not beaten a private. 20 of these have an overall winning playoff record. The other 160 have a cumulative playoff record of 0.271. They aren't really winning against most teams, private or public (478-1014 against public schools and 0-274 against private schools - mostly in small samples). Simply put, there is a early round competitive balance that's generally unrelated to public v private. These schools probably make the playoffs less than half the time, and they're rarely a tough opponent for whoever they're playing.

Overall, including those schools, you have 377 of 524 public schools who are the "have nots", with an overall losing playoff record. Only a handful (about 20) would flip their overall postseason record to > 0.500 if they had never played a private school.

This imbalance dichotomy does exist on the private side, but at a different scale/percentage. Here, you have 29 of 61 schools with a losing playoff record, a little less than half. And obviously since a greater portion of their total games will be against publics, their public v overall record tracks closely.

But drilling down even further, you have 11 of the 61 privates driving an overwhelming amount of the success: Driscoll, Mt. Carmel, St. Rita, IC, JCA, Naz, Montini, Providence, SHG, Newman, and Loyola account for 435 of the 879 private verse public wins at a win% >70%. And they make up half of all private school wins (beating other privates at a >50% share as well).
(Edit/note. This is 61 who made the playoffs in that time span. Doesn't include those who made no PO appearance in 21 years, if any - though I really didn't catch any obvious missing from the list)

I could also add the next three best win % private schools in Marist, Boylan, and Marian Woodstock. At that point, the remaining private schools now compete at a 0.500 level against publics. So about 1/5 of the private schools are really where this discrepancy shows. One is closed down, 2 others have only been juggernauts for about half this time frame.

Only 8 other private schools after these 14 advanced past a quarterfinal (7 having won 1 title and 1 having won 2). Though it should be noted neither Marian or Marist has won a title in this timeframe, placing 2nd once and twice, respectively). In total 53 1st and 24 2nd place trophies from the top 14 and then 9 / 4 from the group of 8.

Now if I took the top 20% of publics the bottom half of that list wouldn't be nearly as impressive at the bottom, but I could probably build a really strong 20-30 public schools whose success would look a lot like the private school 14. In fact, I can create a list of 36 who rank in the top 50 (among public schools) in both total playoff wins and playoff win %. They account for 69 first place trophies and 58 second place trophies.

Your next tier of Public schools (20 next best combined total wins and W%) account for another 10 and 24 1st/2nd trophies.

Now with ALL of this, of course, the trophies are split amongst classes because we first assign a classification and then the cream rises to the top. I wouldn't suggest that St. Theresa be elevated into a have group with Loyola anymore than I would suggest Forreston and Lincoln Way East be grouped. Whatever happens you're going to have a situation where the best see a disproportionate share of the success. Even if you can "solve" the "private school problem", the displacement created and new opportunity is just so little, because the private school share of the population is so little. High school football, like the world, is very top heavy with the haves.

	% Schools	1st Place Trophies	% 1st Place Trophies	2nd Place Trophies	% 2nd Place Trophies	% Total Trophies	Trophy Return Multiple
Top 14 Privates	2.39%	53	31.55%	24	14.29%	23.92%	9.6x
Top 36 Publics	6.15%	69	41.07%	58	34.52%	37.80%	6.1x
Next 8 Privates	1.37%	9	5.36%	4	2.38%	3.87%	2.8x
Next 20 Publics	3.42%	10	5.95%	24	14.29%	10.12%	3.0x
Other Publics	80.00%	27	16.07%	58	34.52%	25.30%	0.30x
Other Privates	6.67%	0	0.00%	0	0.00%	0%	0.0x

I'll pick on Antioch Twitter guys specifically a little here as well. In years where they made it past the first round:
04-05: +21 margin against public , -29 margin against public
08-09: +62, +1, +1 (publics) and -24 (public) in Semis
16-17: +1 (public) and -33 (public)
18-19: +49 and +1 (publics) and -7 (private) in quarters
19-20: +57 (public) and -7 (public)

Their other 2 losses against privates in Rd 1 (17-18 and 22-23) were by 22 and 21. Granted with some small, imperfect sample sizes, but... it doesn't look like the private schools are a competitive problem anymore than other publics are. Sure the draws can be super unlucky and random... but everyone deals with that. I'm sure its salty being eliminated early when you probably weren't a bottom 16 team in your class... but that's the imperfect seeding IHSA has.

This topic once again seems to be a topic of interest, so I am bringing this thread once again to the forefront. The most important part of the discussion begins in earnest about 3/4 of the way through page five. It is there, in the second paragraph of his December 16, 2022 post, that "Snetsrak61" states private schools have won 65% of their playoff games with public schools since 2001. As one can see, the statement was based on extensive research.

 

[crazylegs777]

They split Public's & Privates where I live & I think it's stupid.You get a public school wins a state title then inevitably say oh we could have beat the private state champion.But then if they played & the private school won it'd be they cheat they recruit 😒😒😒

 

[crazylegs777]

Phil Acton said it best.Do i like the fact i'm playing against kids in my own town that live across the street from my high school? No.But no excuses.We just have to get better & beat them!!!

 

[cigaros]

How many private schools are there? I wholeheartedly agree with you, but maybe, the angry mob is upset because of the ratio?

I think you are precisely correct. People are more worried about equity rather than equality. And in doing so, will dilute the quality of matchups and actual champions.

 

[juschill]

How many private schools are there? I wholeheartedly agree with you, but maybe, the angry mob is upset because of the ratio?

Private schools are less than 1/3 of the schools in the state.

 

[cigaros]

State Series Information & Results | Football | IHSA Sports & Activities

www.ihsa.org

513 schools had programs in 2019, the last year of the available data.

 

[Snetsrak61]

There's also something not sitting well with Antiochs methodology though I can't quite put my finger on it. In a single elimination setup with a pretty big disparity in types of schools (public v private) I don't think you can compare head to head like that. It's likely a fractional amount of matchups, but since public v public v private matchups are excluded you get potentially weird circumstances in the public v private matchups that probably are going to "favor" the minority position if there is any sort of success.

I think I finally figured out why this methodology doesn't sit well with me, and I have an analogy / thought expirment to run through. New post incoming later today....

 

[Snetsrak61]

Alright here's my hypothetical/analogy for why the particular method of head to head analysis wasn't sitting well with me. I'm not saying that certain advantages don't exist. Just that certain shortfalls may exist in the particular macro statistical analysis used, particularly if it's primarily a question posed regarding how to best serve competetive balance interests. Only took a year for me to cone back around to it.

The goal here is to show why a head to head matchup analysis between two demographic groups may not be the best methodology to measure competetive balance in a single elimination series where there is a large participation disparity in the majority v minority demographic group (I.e
public v private), as well as broad and large talent disparitys across the entire participation group.

It's going to step outside of sports. And it's of course just an analogy, along with a hypothetical, but I think a reasonable one given natural constraints of hypotheticals and analogies. So please save the nitpick for why it's not a 1:1 analogy. Please. Or arguing the hypothetical. Please. Again, just looking at the possible correlative/causal logic flaws, and isn't to refute private-public W-L data as presented (referring to my own analysis back on page 5 or Antioch's)

Scenario we are using is Chess. There are way more men chess grandmasters than women, which leads to the question: are men inherently better at chess? The consensus from most on this question is that men aren't inherently better at the skills required for chess, but there is a huge discrepancy in participation rates.

Hypothetical me would like to prove this further. So I set up a state championship of chess. However I know I need to fuel female youth participation so I provide the funding (travel, training guides, facilities, etc - hypothetical me us very rich) for any female participant while any male participant must provide their own funding. But they aren't excluded from participating.

As a result, I have a large discrepancy in M-F participation at a rate of about 7 males to 57 females in a typical 64 person state series tournament. As I've ran my state series tournament over the years an observation is made though. Each year, there's about 5-8 true contenders, of which the M-F split is about equal to slightly favoring the male particpants. So let's say in a typical year I have about 3 females and 4 males who route the rest of the competition and have a pretty easy path to the quarterfinals unless they get unlucky and face each other in the first three rounds. After quarterfinals, these groups fare very well competetive against each other.

As a result, when you run the numbers, not only are males disproportionately winning trophies (about 27-35% of all 1st and 2nd place finishes), they also win about 65% of all matchups against females. All this despite being only about 10-12% of the total tournament competitors.

So have I proven that males ARE disproportionately more prone to success at chess? Should the males be placed into a separate tournament? Or is something else going on?

I'd counter that I've created a tournament with unequal participation that creates weird percentage results. And I have a competition with stark talent differences between the best players and and "base players". But my total M-F "top competitor" rates ARE close to 50/50, to match the overall population. The slight male edge perhaps explained by the same macro participation issues that are outside the control of my own tournament and my own attempt to close the gap by addressing participation rates.

I've effectively created a imbalance of the bottom tier, where bottom ranked competetors of one class have low cost of entry to keep playing and bottom ranked competitors of the 'minority' class are probably incentivized to just drop out and compete in other endeavors. So they just don't participate at all. Thus, macro numbers based on participation are skewed. But isn't indicative of an inherent unfairness between both groups.

Now, even taking my hypothetical analogy at total face value, this isn't to say that's what's happening in IHSA football. But it is one possible explanation for a possible logical gap in the type of analysis used when (1) stark competetive gaps in talent and (2) a very unbalanced participation rate between two groups. And it's possible I could choose any identifying quality (teams who's mascot aren't mammals, birds, or warriors!?) and create a similar correlation on a minority group who make up 10-20% of the total competetion.

I could try and solve for the one correlative trait by moving all of one group out (in case of IHSA, move all private schools out to address the 20% or so if them that are a tier above). But I'd almost certainly have the net result of just further reducing the state series results as the best competetiors of the other demo (top public programs) just continue to route their way through lower ranked teans with even less competition left to stop them. Will that be a better state series environment that I left behind? Hard to argue, IMO. But I'm also not the public school coach who now gets to lose out in the semis instead of the 2nd round or quarters. As my prior breakout showed, we don't have to move that far down the private-public school success rankings until all of the competetive imbalance results are totally washed out between public and private schools. At the top end, the top private schools haves marginally more success than top public schools, but even if we want to address that marginal imbalance, and narrowly address it, I think we should admit that hyper focus on one difference (boundary v non-boundary) is unlikely to sufficient address the entire body of participants where really 20% or less of the participants rule over the rest of the competetion in either case.

 

[Teetime]

Knock yourself out.

Why use Loyola as the private school example in this study of college rosters, but exclude ESL from the same sort of study? Compare the best to the best.