At what point do you guys bail on a player?

This is my starting PF in DII (motion offence):

Athleticism 52

Speed 77

Rebounding 97

Defense 50

Shot Blocking 73

Low-post 88

Perimeter 42

Ball Handling 13

Passing 15

Work Ethic 60

Stamina 81

Durability 36

FT Shooting D-

IQ A

To me, this guy should be a stud. He has phenomenal speed for his position, has a great low post rating, as well as an excellent perimeter rating for a DII big. He does have BH/PA issues, but I usually try and play him with my 44 passing center.

However, his results against zone defenses this year (12 games) are 7.4 PPG on .356 shooting. Should I be killing his distro, since the results are the results, or should I trust the ratings, which haven't exactly been playing to form (see: 6.6 REB for a 97 rated rebounder).

At what point do you guys bail on the ratings and go by the results? And is there a reason in the ratings that I'm not seeing as to why he would struggle so badly against zones (in 16 games against full court presses and man defenses he is scoring 8.9 PPG on .536 shooting FWIW).

Thanks,

homr

i would say never completely bail on ratings. but sometimes you just have to accept something is going on that you aren't accounting for (could be some nuance in the sim engine that none of us will ever discover). is 12 games enough? not really. but i would probably take it as a decent indicator, and start decreasing distro marginally.

i am a huge proponent of playing based on a player's stats, but they are *extremely* subjective. so you have to be really careful - thats why i say never completely bail on the ratings. hence my suggestion to decrease marginally, not "kill his distro" :)

also, sometimes a simple lineup switch makes all the difference, for no apparent reason. maybe its just luck, but i doubt it. usually when a guy of mine badly underperforms, i move my depth chart around a bit and see if it continues.

oh, sorry, i read the end but apparently was thinking about something else and completely missed the fact he was doing fine against other defenses. well, in terms of interior defense, i would think the 2-3 zone is about your best bet, as it is the worst for perimeter defense. so, that might be a part of why you are seeing such a discrepancy. i would also say randomization is a factor, i doubt his steady state performance would be so drastically different, unless there are factors out of the discussion (like, were the zone teams better on average than the press/man teams? were their bigs better? what about the +/- settings, my impression is zone teams play - defenses as much if not more than any other defense, so maybe that is part of it)

Quote: Originally posted by coach_billyg on 3/10/2010i am a huge proponent of playing based on a player's stats, but they are *extremely* subjective. so you have to be really careful - thats why i say never completely bail on the ratings. hence my suggestion to decrease marginally, not "kill his distro" :)

This is a big part of the reason I ask this question; every season I track my players against each of the defences to see if trends develop (ie a bench player doing well against a certain defence, you can utilize later in the season to surprise an opponent). What I've found though, is that one's seasons results vs certain defences doesn't carry over to the next. I've had guys with heavy distro shoot over 60% against presses, only to fall to under 40% the following season.

do you track your SOS against those defenses as well? usually when people talk about such drastic performance gaps from one season to the next, SOS has changed a lot in the way you would expect.

with backups, i would just suggest a little caution on the stats, sample size is such a factor, actual stats are often meaningless. especially if you have a lot of blowouts.

I never thought of SOS; that is a great idea, especially in a conference that is half-sim. How do you calculate that though? Just the winning percentage of the teams?

well, that is an approximation, but not correct. SOS is actually the 75% of the RPI equation that is not your winning percentage. so while RPI is
25% * (your weighted win %) + 50% * (your opponents weighted win %) + 25% * (your opponents opponents weighted win%),
SOS is
.667 * (your opponents weighted win%) + .333 * (your opponents opponents weighted win%)

in all cases, games against a team do not count in the calculation of their opponent's win %. so if you play team X, in the 50% of your rpi formula, their winning % does not include that game. however, you can be included in your opponents opponents weighted win% part of the formula (you play team X, who plays team Y, who also played you). hope that makes sense.

now, it is a bit difficult to calculate that without much effort, or a program to gather the data. so i would just make something up based on opponents rpi and win %.

an often overlooked conclusion of the above is scheduling home vs away does NOT affect your SOS. if you play 10 teams in non conf, your SOS is the same whether it was 10 home games or 10 away games. so SOS in the way it is defined is really not that good of a measure of "true strength of schedule" if you will, in my opinion. that said, i bet a half-assed approximation based on rpi and winning % is probably just as good :)

He has 2-10 and 2-16 games in there against zones. In looking at those opponents quickly, the starters have good block and defense ratings and likely were big reasons in why he had an off game. Before every game, I take a look at the guys my top scorer(s) go up against. If they look like tough defenders, then I adjust accordingly and try to find a mismatch to give distro to. GL!

Quote: Originally posted by jcf16 on 3/10/2010He has 2-10 and 2-16 games in there against zones. In looking at those opponents quickly, the starters have good block and defense ratings and likely were big reasons in why he had an off game. Before every game, I take a look at the guys my top scorer(s) go up against. If they look like tough defenders, then I adjust accordingly and try to find a mismatch to give distro to. GL!

not a bad thing to do. but keep in mind the sim engine is already taking into account how open your guys are in deciding who takes the shots. basically, they are already adjusting for you. making further adjustments can be beneficial, but if you are tweaking by more than 1-2 on any sort of regular basis, you are likely overcompensating and hurting yourself. occasionally an extreme man defense mismatch (taking into account not only the other starter, but everyone else who will defend a player of yours), i would make more extreme adjustments. but, very rarely is that the case.

Quote: Originally Posted By homrbush on 3/10/2010

Quote: Originally posted by coach_billyg on 3/10/2010 i am a huge proponent of playing based on a player's stats, but they are *extremely* subjective. so you have to be really careful - thats why i say never completely bail on the ratings. hence my suggestion to decrease marginally, not "kill his distro" :)

This is a big part of the reason I ask this question; every season I track my players against each of the defences to see if trends develop (ie a bench player doing well against a certain defence, you can utilize later in the season to surprise an opponent). What I've found though, is that one's seasons results vs certain defences doesn't carry over to the next. I've had guys with heavy distro shoot over 60% against presses, only to fall to under 40% the following season

homr - this kind of crap drives me batty sometimes. Here is a suggestion that has worked for me sometimes. Switch lineups or fatigue settings so that he is playing with different players. If you go through your PBP you may find that he is often stuck on the court with weaker players and this causes him to underperform. Here is a simple lineup example. Assume your starting 5 is strong but your backups are weak.

Starting Stud PF in question - 99 STA ; All other starters - 80 STA. What frequently happens here is the other starters sub out while the stud PF is still on the court. He becomes the focus of the defense and then underperforms.

Just a thought. But sometimes I will make unconventional lineup changes that end up working.

mully, i think that is great advice. WIS claims a guy's ability to make a basket has absolutely nothing to do with his team mates, what so ever, in any way, directly or indirectly. once he has the ball, doesn't matter if the other 4 guys are all 100s or all 1s. but, i think we all have to submit to reason and experience on that one, and figure WIS is off the mark. its seble who says that, so i take away that the effect is not obvious, but very much real. like, the defense focusing on the better guys, or something to that effect (instead of the equation directly taking into account, for example, the passing ability of the point guard).

Quote: Originally Posted By homrbush on 3/10/2010

Quote: Originally posted by coach_billyg on 3/10/2010 i am a huge proponent of playing based on a player's stats, but they are *extremely* subjective. so you have to be really careful - thats why i say never completely bail on the ratings. hence my suggestion to decrease marginally, not "kill his distro" :)

This is a big part of the reason I ask this question; every season I track my players against each of the defences to see if trends develop (ie a bench player doing well against a certain defence, you can utilize later in the season to surprise an opponent). What I've found though, is that one's seasons results vs certain defences doesn't carry over to the next. I've had guys with heavy distro shoot over 60% against presses, only to fall to under 40% the following season.

Yeah, that doesn't surprise me. The sample size that you're working with in a season vs. a particular defense just isn't large enough to be reliable.

If you could build a bigger database over a number of seasons (i.e. fast guards w. low pe vs. certain defenses; big men with high pe vs. certain defenses), you'd almost certainly reach the point where trends would develop. That's a ton of work for not very much benefit, imo.

As for his results against the zone, it's still a small sample size, but also not surprising given that 2-3 zone provides the best interior defense.