|
Post by methuselah on Jan 23, 2015 7:27:52 GMT -6
I've seen this topic broached in threads on other subjects and always found it to be a pretty interesting discussion but I think it's been quite a while since it's had it's own thread.
So, here's my admittedly math challenged take on it. Feel free to point out where I go wrong and also to join in the discussion.
It seems to me that a lot depends on what goals your team is (realistically) looking at. And I figured one way to look at it would be to identify a couple of likely goals and look at what power ranking it would take to be in the hunt for them.
The D-I numbers are a little bit higher so I'll concentrate on D-II & D-III.
If you are trying to achieve the holy grail of a #1 ranking you would need upwards of a 13. So for any individual game you schedule, you would need either a win (5 points) against an opponent with at least 8 wins (with an opponent's tie counting for 1/2 a win of course), or a draw (2.5 points) against an opponent with at least 14 wins (if I'm remember right and a draw is .75 of opponents' wins) or a loss against a team with 26 wins (I don't think there are any of those out there).
If you are shooting for top 16 and a first round home game, you would need about 8 points to keep you on track. That would be a win against any opponent with at least 3 wins, a draw against any opponent with at least 6 wins or a loss against a team with at least 16 wins.
If you are just trying to get into the playoffs would need 6 points (5 would do it in D-II but I think that's a one year thing because of how many teams are making the playoffs in comparison with how many teams are in the division). That means any win where your opponent has at least one win would be beneficial. Or any draw where your opponent has at least 5 wins. Or any loss where your opponent has at least 12 wins.
Anyway, that's my probably wildly inaccurate take on it. Any corrections welcomed. What it means to scheduling in the real one way or another I'll leave for discussions for those more knowledgeable than me.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 23, 2015 11:46:54 GMT -6
Good points.
And I think there's a lot of merit to it.
Looking at D1, let's face it, the perennial top 4-5 teams don't change much. Most coaches are going to seek out a mixture of 'training games' and challenging games. The right mixture just depends on the individual coach and the talent he has coupled with expectations.
Safe to say that St. Paul's, Jesuit and Acadiana go into every season driving for a state championship. That's what a winning tradition and a steady supply of talent will do for you.
For most everyone else, I think most coaches aim either for a winning season, a district championship or a playoff berth.
Which means a lot of coaches will try to schedule as many games as possible to win. Whether it's ego driven or parent/player expectations.
Personally, I'd like to get away from formal districts. Some districts are more competitive than others. But nearly every district has at least one, maybe two (or more) weaker programs that really need to play a different level to develop.
That doesn't mean I'm in favor of creating something like we have with gridiron football. because it diminishes/dilutes the quality/meaning of actually winning.
So, back on point, I like the idea of the top 24 teams based on power ratings (based on 3 divisions) making the playoffs. A system that rewards results and strength of schedule.
Under such a system, it would negate the need of formal districts.
|
|
|
Post by kevin on Jan 23, 2015 16:11:20 GMT -6
Keep in mind that all the PR numbers rise over the course of the season. So 14.00 now is better than 14.00 next week.
My philosophy is that there are four types of teams:
1) Good teams that win a lot of games. 2) Good teams that don't win a lot of games. 3) Bad teams that win a lot of games. 4) Bad teams that don't win a lot of games.
Teams in category 1 are the powerhouses.
Teams in category 2 fall into a few groups. Some may not play a lot of games due to early season conflicts with fall sports. Or their coach may not want to go to tournaments. Or they may play a very tough schedule.
Teams in category 3 are teams that may be at the top of a weak district, or who play a very soft non-district schedule. They pile up wins against very weak teams.
Teams in category 4 are the weakest teams.
You really want to play the teams in category 3. Playing the teams in category 1 won't hurt you too much, because they win so many games. But playing the teams in category 2 may be a bad idea, especially if they're strong enough to beat you. And playing the teams in category 4 won't help much, either,
Unfortunately, you don't always know which team will fall into which category. But generally speaking, you want to play against teams that play a lot of games. If you know that a team plays an out-of-state tournament each year but only a bare minimum schedule in Louisiana, you might not want to play them.
|
|
warrior16
Data Expert
Michael Stein - Volunteer Assitant
Posts: 2,169
|
Post by warrior16 on Jan 23, 2015 16:40:49 GMT -6
When we made our non-district schedule for this year, we did it with power rankings in mind, but our non-district schedule remains almost completely unchanged from last year. We thought adding three tough road games against Baton Rouge, Dutchtown, and East Jefferson would make us a better team in the long run, as well as playing teams like McKinley, St. Amant, and Denham Springs. We knew there would be a very high possibility we would lose most of these games, but I'd rather lose a few games on the road to good D-I teams than coast against smaller schools and not get tested, which we could have chosen to do.
Much better to play a brutal schedule and earn a #4 or #5 seed than to play an easy schedule and get the #2 seed, in my opinion. The team that tested themselves more during the regular season is far more likely to win the games that matter the most than the team that did not.
It is theoretically possible to "game the system" with clever scheduling even with all the variables involved (not that I'm saying any team is doing that), but that won't help in the playoffs when you have to play a good team.
But Hall is right. For some teams, just getting to the playoffs is the goal for the season, while other teams have bigger goals. You get out of your schedule what you put into it.
|
|
|
Post by coachray40 on Jan 23, 2015 16:44:53 GMT -6
I beleive the total # of games scheduled makes a difference too. Schedule too few and each win and loss swings your PR wildly. Schedule too many and the amount of points it takes to keep your PR up gets higher and makes your upswings smaller. I would suggest no more that a total of 18 games--that seems to me to be about correct.
As far as teams go, you dont want to schedule any team that doesnt win a lot and could end up beating you. Either schedule better "lock" teams that you are sure to get a minimum of 5 points from, or play teams with a lot of wins that give you plenty of continuous point growth, even if you lose. Great teams to have this year are Denham Springs (probably 24 wins), University (20-22 wins), St Louis (20 wins), Episcopal of Acadiana (lot o wins and ties).
Even more importantly, play teams that dont lose much. Teams that win AND tie help. The big thing is for your team to be able to gain down stream points after you have played them, and ties help this process too. Baton Rouge High is a good team with ties right now as I think they have 7. Thats seven games they didnt lose.
Look for that ONE game that will put you over the top. Episcopal scored big with a win over Parkview. Ultimately that win will probably give Episcopal 20 points. That is a full power rating poin across a 20 game schedule--for only one result. Makes the rest easy.
Finally, give your schedule half your toughest opponents at the beginning, and half at the end. Get some points early and keep growing your PR, whil having teams that can give you at least 10 points even if you lose at the end. You want to be trending up in power rating at the end of the season. Im keeping a detailed spread sheet of the Power rating s for the entire season, and its very interesting to see some team surging at the end of the season (St Louis, Christian Life) and other teams fading.
|
|
|
Post by beauchenecoach on Jan 24, 2015 13:38:44 GMT -6
You schedule based on your needs and goals. Want to make playoffs onky, schedule a bunch of weaker teams you can beat... But you'll be a weak seed. Want to be a home team, schedule 10 game wining teams you can beat. Want the top seed, schedule as many 15-20 (closer to 20) game winners as you can. Then go earn it by beating them. We have played one of the most complete Division II and III schedules and lost one game and tied 5. St Louis did the same in DIII. That's why we both are where we are. We have increased our gap over the last couple weeks playing teams that actually were lower than our average in district. Why? Value of our schedule before district. Those teams are all winning 1-2 games per week... Increasing our score each week. our district has great teams in it despite some of their records. Our district is way harder than its record. I challenge any district in DII to let us know which district is harder than ours despite the records this year? Would you trade your 2nd and third place teams for STM ad Teurlings twice on your schedule? Our last place team (Opelousas) tied the third place team on BR district (Plaquemine) and beat other districts third place teams. Now I know, BF and EJ are a great top 2 in their district but besides them and that district? EJ also played a great power rankings difficult non district schedule and won to keep themselves in the top 4-5 when it's said and done. Another win here or there and they could have possibly been a top seed. You still can't forget... The schedule sets up your goals, but you must have success in that schedule to achieve the goal. If 2-3 teams in your district do this, it will raise the bottom half too. When we played Opelousas earlier in the year, they jumped 5 spots after losing to us. Let's see if the same happens next week when we play them. Or look at Westgate in the next ratings. The lower teams get boosts from playing the top teams in their districts if those top teams scheduled right and are having success. Back to our schedule... we figured we would need a huge boost outside of district cause we still play each twice... Something that won't happen in future years. I think the days of playing twice in district are over with in all areas except the areas that are travel restricted and out of the way. Teams need to be in more control of their destiny and goals with scheduling. In 2 weeks, all of the numbers will be final and will also be 2-3 Pts higher. And I also think, with a few exceptions, the PR will closer match the coaches poll at the end. Maybe a spot or two off for a few teams. Maybe one anomaly in each division. But like mentioned above...a spot or two off could make a game that should be at least a Semfinal matchup, happen in the quarters instead. That's the one thing I wouldn't mind see changing.. Add a final step to seeding where coaches can seed but can onky move a team a maximum of 3 spots from its computer ranking. This would eliminate any attempt to play with the seeds by a human but also allow coaches to fix what is widely seen as a wrong seed, either too low or too high.
|
|
|
Post by rora on Jan 24, 2015 13:48:24 GMT -6
Didn't we have a formula which took into account opponents' opponents? Seems to me an important factor omitted in the current, to me, simplistic formula.
Thoughts on this? And would the LHSAA ever adopt a formula like it?
|
|
p_malinich
Data Expert
www.elevenlions.com
Posts: 4,201
|
Post by p_malinich on Jan 24, 2015 13:52:13 GMT -6
Didn't we have a formula which took into account opponents' opponents? Seems to me an important factor omitted in the current, to me, simplistic formula. Thoughts on this? And would the LHSAA ever adopt a formula like it? We've never tested any models like that (which would be more complex). There may have been discussions suggesting it. I think that's how the RPI is calculated for NCAA purposes.
|
|
|
Post by beauchenecoach on Jan 24, 2015 13:53:54 GMT -6
Didn't we have a formula which took into account opponents' opponents? Seems to me an important factor omitted in the current, to me, simplistic formula. Thoughts on this? And would the LHSAA ever adopt a formula like it? Without being negative to any association... No they could not. Our simple formula this year Took weeks to get correct. There just isn't a true understanding of our sport yet.
|
|
warrior16
Data Expert
Michael Stein - Volunteer Assitant
Posts: 2,169
|
Post by warrior16 on Jan 24, 2015 14:04:38 GMT -6
Yeah. RPI is an incredibly complicated metric to put into the computer, and no high school association would have the manpower to operate a system like that. Not only do your opponents' results factor into the ranking, but your opponents' opponents' results factor into the ranking, and every change in the schedule changes the metric for all teams involved.
I'm not sure if the NCAA uses it for soccer, but I know they use RPI metrics in basketball and baseball (there was lots of controversy in 2011 when LSU got left out of the baseball tourney with an RPI of 26). In college schedules are made and confirmed sometimes years in advance, and we don't have that kind of luxury in high school, especially when some games will rain out.
Maybe one day we can use a system like this, but not for the forseeable future.
|
|
|
Post by rora on Jan 24, 2015 14:07:40 GMT -6
Didn't we have a formula which took into account opponents' opponents? Seems to me an important factor omitted in the current, to me, simplistic formula. Thoughts on this? And would the LHSAA ever adopt a formula like it? We've never tested any models like that (which would be more complex). There may have been discussions suggesting it. I think that's how the RPI is calculated for NCAA purposes. Thank you. Could not remember exactly. Any statistics majors out there care to devise one for fun which takes into account opponents' opponents? With three children age 7 and under, my statistical formula for free time approaches zero.
|
|
|
Post by rora on Jan 24, 2015 14:14:23 GMT -6
Didn't we have a formula which took into account opponents' opponents? Seems to me an important factor omitted in the current, to me, simplistic formula. Thoughts on this? And would the LHSAA ever adopt a formula like it? Without being negative to any association... No they could not. Our simple formula this year Took weeks to get correct. There just isn't a true understanding of our sport yet. I do think it needs to remain simple in operation for that exact reason. And as warrior16 says, that my not be possible.
|
|
|
Post by rlb2024 on Jan 24, 2015 14:43:08 GMT -6
So when the season is all finished how many points do you think will be needed to be in the final 32 in each division? In the top 16?
In Division I there doesn't look to be any danger of a district champion finishing below 32 in the final power rating, as the lowest district leader is Ruston at 16 as of this week. Same with Division II with Lakeshore at 13. Division III could be iffy, with Country Day currently at 33.
This is all a guessing game with no historical data to go by, which makes it really interesting.
|
|
|
Post by coachray40 on Jan 24, 2015 17:49:04 GMT -6
If the season ended today Country day would by 34th--I have a spread sheet that I am keeping to watch the race for the final spot.
I wouldnt worry about Country Day though. Their remaining schedule will be more than enough to get them in, with games against Northlake, Haynes, Houma Christian, Lusher and Catholic-NI. I only really see one game I would favor them in (a 6 win Houma Christian team) so I would expect them to get mad points even if they lose 4 out of the last 5 and still be above the cutoff.
Right now West Feliciana is on the hot seat at #32 and a downward trend in power ratings--they have dropped 1.06 points across their last 5 games. As of today they are holding a 6.26 power rating, but with games remaining against Runnels, Walker, and a 0 wins Catholic PC, I beleive they will stay on the downward trend, even if they win their last 2.
Two other teams fighting to stay in are Riverside and Hannan. Both had good starts, but recent results have not been helpful. Riverside has (today) a 6.73 PR, with games left against a suddenly vulnerable Parkview, University, AC, Dunham (who they lost to in the first round of district play) and Lee. Keeping up that 6.73 will be tough, with most likely little to no upswing without a win against either Parkview or Dunham. Lee and AC wont help, and nobody is beating University so it will be key for Riverside to get at least one W vs Dunham or Parkview. Those games are huge. Hannan has now dropped to #30 as of today (at 5:38 PM) and has lost .77 in power rating over their last 5 games. Only two games left are Pope John II and St Thomas Aquinas, both of whom Hannan lost to first time around and neither with more than 6 wins. Losses to each will offset power point gains from earlier games against Northlake and Catholic-NI, but with nine earlier matches against teams with an overall record of 12-101-9, Hannan isnt getting much help. Without a win in those last two games, I see Hannan being in real trouble on 2/3
|
|
|
Post by coachray40 on Jan 24, 2015 18:21:25 GMT -6
Ad just that quickly it can change. By virtue of the Newman, Westminster, Northlake and Episcopal of Acadiana results, West Fel now has been replaced at 32 by.......you guessed it--COUNTRY DAY. The changes are happening very fast now as teams are moving up and down the board as results come in.
The big winner today was Dunham who scored an 18 point result by beating Parkview. This morning Dunham was at #33, now they sit at 27. Fun stuff
|
|
jk52
All-District
Posts: 216
|
Post by jk52 on Jan 24, 2015 18:28:24 GMT -6
Does a team ranked #23 (Notre Dame)with less than 15 games get into the playoffs ahead of No. 33 (West Feliciana)?
|
|
p_malinich
Data Expert
www.elevenlions.com
Posts: 4,201
|
Post by p_malinich on Jan 24, 2015 18:29:38 GMT -6
If the season ended today Country day would by 34th--I have a spread sheet that I am keeping to watch the race for the final spot. I wouldnt worry about Country Day though. Their remaining schedule will be more than enough to get them in, with games against Northlake, Haynes, Houma Christian, Lusher and Catholic-NI. I only really see one game I would favor them in (a 6 win Houma Christian team) so I would expect them to get mad points even if they lose 4 out of the last 5 and still be above the cutoff. Country Day has also added a game with Franklin on their schedule. That gives them 4 games this week, but also a 1-game cushion to the minimum 15 games. However, it shows Catholic NI as "P" (postponed) on 1/31 so maybe that one isn't happening & Franklin becomes their 15th game.
|
|
|
Post by beauchenecoach on Jan 24, 2015 18:30:21 GMT -6
Does a team ranked #23 (Notre Dame)with less than 15 games get into the playoffs ahead of No. 33 (West Feliciana)? Yes, but will be seeded behind all teams who have played 15 or more games... So probably 32
|
|
p_malinich
Data Expert
www.elevenlions.com
Posts: 4,201
|
Post by p_malinich on Jan 24, 2015 18:31:21 GMT -6
Does a team ranked #23 (Notre Dame)with less than 15 games get into the playoffs ahead of No. 33 (West Feliciana)? The #23 team would drop to bottom of list if they don't have 15 games. However, if they win their District, then they only drop to the bottom of the playoff teams. I don't think they will so they would drop to the 50's and #33 would move into #32.
|
|
|
Post by beauchenecoach on Jan 24, 2015 18:34:27 GMT -6
Does a team ranked #23 (Notre Dame)with less than 15 games get into the playoffs ahead of No. 33 (West Feliciana)? The #23 team would drop to bottom of list if they don't have 15 games. However, if they win their District, then they only drop to the bottom of the playoff teams. I don't think they will so they would drop to the 50's and #33 would move into #32. No, this wrong. (sorry Paul! ) They are in playoffs if they are in top 32... They just drop to behind all those who have 15 or more games.
|
|