Posted by gillispie1 on 11/22/2020 3:00:00 PM (view original):

Posted by npb7768 on 11/22/2020 12:39:00 PM (view original):
Also, i just read the thread on EE started by Mully, but i got lost in Gillespie's thorough posts... LP and PER may be important, depending on the EE's position ?

i don't even know what i was talking about so you are good. i would not call those posts thorough but i appreciate it nonetheless :)

short version - holding back the gaudy ratings via practice planning you don't need to stop your guy from going over 'key points' on the board.

long version -

first, some facts:
1) player stats and playing time do not affect big board rankings. while all american awards and perhaps the NPOY impact final draft rankings (the big board we don't see), as to NT success, there is in general no reason to hold back players performance wise, only ratings wise (although i've held back guys who were pushing into national AA award range on distro a bit to avoid a penalty there - i mean, its intended as a boost - but obviously i don't want EEs so i see it as a penalty). note: conf awards are irrelevant.

2) the big board uses different equations to evaluate different recruits. i used to call these position rankings as a way of trying to tie the concept to something basketball-y, but i don't think the way HD EE formulas work is close enough to what most of us think of as positions, to really tie the two together. long story short, a guard and big follow different formulas. this isn't terribly important for new-to-EE-manipulation folks. but as you get into it, its good to know.

3) the way the big board works SEEMS to be simple linear formulas. every additional 1 point of per might give a player 1 extra point towards the big board rankings. suppose this is a guard, maybe passing is also 1 point each while rebounding is 0.2 points each (just made up numbers). its not 1 linear formula for the whole board, because of #2, but still - that is the way i'd think of the board.

4) class does NOT tie into big board ranking. it only ties into odds of leaving.

5) the way the draft works is it basically goes to the first guy on the board, and says hey, if you were going to be the #1 overall draft pick, would you leave early? the player goes, hell yeah i would! well, unless my mom, dad, and first cousin all die on the same day in a horrific accident. then i will be too grief stricken to accept nba money. so long story short - 99% chance yes. the game goes ok, let's roll for that - you didn't hit the 1%, you leave early. then it goes to player 2, who might have a 98% chance of leaving early if he was going to be the 2nd overall pick. the draft proceeds like this, player by player, until the draft is full. note that for graduating players, they always are 100% to take the draft if offered, while for underclassmen, their class substantially affects their odds of taking the deal (with younger players staying more).

6) the main thing that changes the big board to the final draft board is NT success. if you win a title, expect your players to jump about 10 spots (in the important ranges - it will be less at the top, and perhaps more at the bottom). title chasing teams should keep this in mind.

ok, now that that is out of the way. basically, when you have top tier players, figure out the crap they don't need, and don't grow it. that is roughly what it comes down to. but really you want to incorporate this not just into practice planning for an on-the-board junior. you can try to affect the outcome then, but often, its going to be too late. you want to work on projecting out where your players will fall based on their ratings, getting a sense of which ratings like lp and per are really valuable big board wise, and when the player might not need that rating enough to justify getting it, given the EE risk. for example, a guard with 95 ath/spd/def/per/bh and 75 pass is going to be a risk. let's say hes 15 lp green. i'm going to think, hmm, at a +2, this guy is amazing and i don't really care to grow his 2pt% by a little bit when 80% of his shots are 3s. i'll leave him at 15 lp green *from freshman year*. now junior year, i might reconsider now that i see where my guy is on the board. and senior year, i probably am putting 50 minutes into that lp because i have nothing else to do anyway (although if i am desperate for 3s as a team, i wouldn't even take that lp for free, so then i wouldn't - but that's not really relevant).

ideally, you carry this back even further than freshman year - into recruiting. ideally you've already projected out a player's EE risk and know how you'd handle it, before you seriously recruit them. usually i make this determination for some recruits before the first cycle and then for other ones later, as the situation crystalizes. you don't have to get too crazy about this. i can make that determination in about 30 seconds. actually what you REALLY want to do, is before recruiting a player, have figured out all of: their EE plan, their player development plan, and how this player will look NT time each season and what role they will play (backup sg, medium scorer, as a soph guard - or perhaps starting big, key scorer as a junior big - etc). those things all go together, there is no order - the EE risk affects the player development plan, but the player development plan also affects the EE risk. they all feed into each other. once you have a nice, stable view of how this player will look, grow, and perform over the course of their entire career, NOW you can actually evaluate how valuable the guy is (to make life simple, 90% of the time you can just think - what roles do i need this guy to fill end of soph and end of junior years, and what does it take to get them there. those are the 2 key optimization points. one of the absolutely most consistent differentiators between the absolute best programs and the regular a+ ones is how well developed end of season sophs are). and making these evaluations prior to spending your recruiting dollars will allow you to target really high value players for your organization, to really line up your recruiting effort (ap + money) spend with the recruit's value, which is essential to maximizing your recruiting success overall (people all the time reach out to me like, i suck at recruiting, can you help - and its like no - you are actually doing well enough getting the guys you are targeting, you just have no idea who you are supposed to be getting yet. without understanding what makes teams and players good down the road - both generally and specifically for your particular circumstances - and therefore how valuable a potential recruit is to your team - how can you do a good job spending your recruiting money efficiently? you can't. while recruiting is important, team planning is essential for everything, which is why i am always on team... team planning?).

ok, so that is sort of more the long game, and the more you can incorporate EE planning into your team building decisions, ideally from the beginning (recruiting), the better off you'll be.

but really, where are the major gains from all this? half of them are frankly from the planning. in a sense, EE manipulation is merely the gateway drug to actually doing team planning right, and that is where a TON of the benefits come from. but you also can't start there. so what are the benefits of EE manipulation proper - not just EE planning? well, there are key ranges on the board. those ranges are roughly as follows: around the 30 mark, and around the 100 mark. its probably easiest to go about this by class, but note that on the board, 28 and 18 are the magic numbers where the listed ranges change, if you are into that sort of thing, but not necessarily where the inflection points on the odds really are.

graduating players - just get them as high as possible always
juniors - between 1 and 20 or so, these guys are basically gone. they tick from likely going to on the fence when they go from 28 to 29, and are always either 'likely going' or 'on the fence'. you have a low % chance of getting really lucky with a 1-20 junior, but its too low to plan on, too low to worry about. between 20 and 40 on the board, you get a rapid change in odds, dropping from close to 90% down to around 50%. but the odds of a junior leaving never drop that much lower. really, by 50 on the board, i'd say a junior is 35-40% to leave, and that doesn't change meaningfully anywhere from that 40-90 range. so if you have an on-the-board junior at 70, you really may as well just grow him - just don't let him go past 40 or so, maybe 50 if you are one of those top few teams, hoping for a title or at least a f4.

the off the board range in the trickiest. if a lot of players choose to leave early compare to the odds (say of 10 juniors with 40%, 6 leave), then fewer players will even get a shot at getting drafted. some years there's a ton of quality seniors so less underclassmen get a shot, other years its the reverse, and so forth. usually, at least everyone in the top 90 gets a shot at leaving, and often down to around 100. especially now with EE manipulation, its very normal that players as low as 120 or so on the board (invisible to us, but if you get good at projecting players, your spidey sense will tell you when you are close) get a shot. being off the board reduces the odds of a junior, but only because he has less odds of getting asked. any junior when asked to leave for the late 2nd round is about the same odds, right around 35%. if your 105 just-off-the-board (say you saw him drop off with 3 practices to go) ranked junior only has a 50% chance of getting his name called, he only has a 17.5% chance or so of getting drafted that year, because if his name IS called, he's still about 35% to take it.

summary: keep your juniors around 40 if you can, EEs are brutal in 3.0, and 40 board ranked juniors can be *amazing*. they tend to me *much* higher overall value than juniors ranked in the top 20, because they are almost as good (often better) but stay longer, and in 3.0, if you can totally avoid that EE penalty altogether, that is HUGE. also, if you have to decide between an off the board and on the board junior, and you can make that off the board guy pretty unlikely to leave, you can really impact your overall EE count, too. but it is far trickier than the 40 mark which changes much less.

sophmores - 1-18 will be likely going, 19-28 is on the fence, 29-XXX is likely staying. in that top 10-15 range, sophs are almost as likely to go as juniors, or maybe just as likely. everything between 15 and 35-40 on the board REALLY matters for sophs. by 40 or so, 50 for a title winner, sophs are only perhaps as low as 10% and as high as 20% to leave. if they go from near 90% at 15 to 10-20% at 40, you get the idea how valuable that range is. in short, having a soph above 20 or so amounts to HD blasphemy. roughly nobody can ever justify breaking that rule. i don't even break it and i'm the sort of the HD equivalent of that guy from inception who spends all movie doing the exact same **** he just spent 5 minutes making a big speech telling everyone not to do. having a soph above 40 is only really justified in extremely important high end seasons. like if i was about to quit a program and really wanted to win a title on the way out and had a good shot at it, i'd be ok having a soph above 40. or if i was going for the 3rd title in a 3peat or something. regular folks can substitute winning a d1 title or their first d1 final four or something. but bottom line - you are selling out your future. only do it if you can really justify selling out your future to have 1 guy be better for 1 season.

note that sophmores off the board work just like juniors. if their name is called, their odds are the same as is if they were on the board. also i don't think i said this, but if a sophmore is 100 on the board and gets called, their odds are still that 10-20%, same as around 40-50. it never goes down from there for anybody, until the players start getting out of draft range completely.

freshman: they are only likely going i believe in the top 8, then on the fence 9-18, likely staying 19-28, and are staying from 29 on. if your freshman is ever in those top 28, or top 40 for NT winners (they can get boosted like anyone else), you should more or less go kill yourself. but not really, only metaphorically. i've seen it happen; its baffling, freshman cannot be good enough to justify letting them leave. roughly you need to watch your on-the-board freshman like a hawk. if you are growing stuff today you'd wish you hadn't when your junior was 25 on the board, or your soph - don't be growing it. if the freshman has good work ethic you are probably screwed no matter what, you can't stop a train like that. so start deciding where on the spectrum of badness makes sense for this player. you need to get value from them, that is commensurate with their EE risk. sometimes this is impossible - these are the guys you probably want to pass on in recruiting before you spent your entire budget on them to end in a 3-way coin flip due to the insane eliteness of the player (you know, if you only look at them and drool, instead of actually thinking).

but normally you'll be able to find something. you have to know when your key years are, if next year is everything to you, maybe you push him (but not above staying, except maybe like a couple times in the entire history of HD) and concede him as a sophmore. this is a brutal move but i've done it to win championships. now normally i still won't go into likely going. i'm risking my future but i'm not completely selling it out, 9 times out of 10 (and this was a rare scenario to start with). if sophmore year isn't that important, its a no-brainer. i've literally done NOTHING to grow freshman before if that is what it took to get them to junior year. like where they have 85 work ethic and are starting on the board, its like uhhhhhh if i don't 0 this guy out NOW, on my very first practice of freshman year, hes gone. its ok though. i'll grow him later, in time for when i need him to be good.

really - freshman year is not for players going early. unless you screw up catastrophically or hit a really unlikely edge case. it about seeing the train wreck coming, and stopping it from happening, while there is still sufficient time to change the future.

OK, that aside. most common stuff. bh/pass for high end bigs is automatic. lp for guards depending on their circumstance, passing for scoring 2/3 guards, defense for scoring pgs, those are probably most common. rebounding on guards has less EE value than real life value so its not a great candidate. its hard to summarize here because if you ARENT team building and planning based on abilities, its just so much harder to give the answer. the short answer is, whatever is outside their core ratings given their core ability strengths, of which each player generally has at least 2 (most have exactly 2). a pg can almost never justify holding back bh/pass, but a scoring pg can justify defense and a non scoring pg is usually safe from the get-go, but can axe lp and per. 3pt scorers of all flavors can axe lp. reb/def bigs can skip lp and per, and usually should. a 95 ath/reb/def big with 90 stamina and 90 block if you are a non press team, that player can be insanely valuable on 0 distro. a 4 year guaranteed player who probably is really good at reb and def (the ability, not the rating) from soph year, with great stamina - what more can you want? that guy can be very valuable to the best programs ever, and certainly everyone else. so for players who won't ever be very good at offense, why even grow the lp/per? also WE and i think DUR does not impact big board ranking and FT either does not or minimally impacts big board ranking. stamina is almost never justifiable as a rating to hold back owing as its role as the most common most important marginal rating for any given great to elite player.

also, don't tank the NT for EE purposes... i cannot get behind that :)

that is the gist of it. identify who is a risk, and what it takes to limit that risk. now re-evaluate their usefulness under that paradigm and decide if they are still a worthy recruit. identify which of your players is a risk, and start holding them back from day 1 in the random crap you don't need. pro tip - you should already be doing this for player development purposes 80% of the time (if you don't put more than 20 minutes into blue/green cores 'because of diminishing returns', this is your wake-up call to start doing so). if any of this EE stuff feels like a leap, you aren't doing enough team planning and player development planning yet, it is truly just more of the same.

awesome stuff! Much more detailed version of how I have looked at this:

focus practice on skillz you need, dont chase skillz you dont need

be happy with 88, you may not need 93

when a guy is high on the board, there comes a time when you surrender to his EE plans

when a guy is lower on the board, getting another season can be worth giving up even useful improvements

and hope that dice roll your way!

Benis - how do you know how much they moved if the big board is never updated? (I have a theory but would like to hear your answer first)

Posted by mullycj on 11/24/2020 1:27:00 PM (view original):
Benis - how do you know how much they moved if the big board is never updated? (I have a theory but would like to hear your answer first)

Yeah I'm saying it is updated. Unless I'm a dingus (very possible) and messed it up before..

We will find out for sure in a couple days after Phelan NT is finished.

What's your theory?

i have never taken this anywhere near as far as these guys appear to have (benis and cub). i will have to check out that video! it is funny, some of the things that i shifted a bit on when benis showed some of his data are some of the things i am off now from cub's presentation. so i think i perhaps need to pay a little more attention :) i never did a 'study' on this, i did pay a lot of attention those first couple years the board came out and kinda helped kickstart EE manipulation as a new line of inquiry. but no way i have a detailed enough insight - or memory - to counter whether the junior drop off is 40-50 or 60, whether a player is 40% at 50 on the board or 50%. i basically only have my own anecdotal observations from a 2 key teams for about 2 years. so i am basically planning to get up to speed on what these guys did, make sure i agree with the approach, and then buy into all their conclusions. the broad strokes are all close enough that i'm thinking they are doing it all right.

the end board throws me. i didn't think the board changed after the NT started, i don't know if its always been this way. benis pointed this out a while ago but i didn't really investigate. i always assume we did not see the final board, but i don't get why the board changes at the end if its not for that, and i did confirm he's definitely right about it changing after the NT. but the changes on the final board look fairly small, so i'm not sure if i just over estimated the impact on NT success and stuff, or if there is still a final board we don't see. my recollection is seble said we don't see the final board. but this final board that comes out, i have no idea what that is then.

there was also a comment about the odds for a 8 soph vs an 8 junior, is the 8 junior the same odds or different - those are definitely different. i would have guessed 8 soph and 20 junior weren't the two equal points, but its definitely different odds based on class - the board rank doesn't change based on class - the odds do. however a comment was also made about different odds *ranges* for on the fence etc for a soph vs a junior, and that does not fit my model (if i understand correctly what was said). i expect to concede that after getting up to speed on that video, but that is not how i had things laid out in my head.

long story short - this stuff is super cool - thanks benis and cub for all your work on the subject! its clear my own understanding has been surpassed by others (i think so anyway), i like it.

Posted by Benis on 11/24/2020 2:25:00 PM (view original):

Posted by mullycj on 11/24/2020 1:27:00 PM (view original):
Benis - how do you know how much they moved if the big board is never updated? (I have a theory but would like to hear your answer first)

Yeah I'm saying it is updated. Unless I'm a dingus (very possible) and messed it up before..

We will find out for sure in a couple days after Phelan NT is finished.

What's your theory?

I think the news on the EE draft results come out in the order they are on the final big board. So although you dont know the position of those who stay I think you can figure out the position of the ones who left. Just my hypothesis.

the news? interesting. what about the actual draft results? i assumed the draft results are reflective of the final big board ranking for all parties who took the leap. and i thought the EE order ended up different than the final board, and that is why i thought we didn't see the final board. but i haven't went back to look at that in 3.0 since benis said whatever it was that challenged my view on the subject (final board results or whatever). and also there could be logic where a team maybe pretends to need a big and so a big gets higher or something, i don't know... but i am curious how the actual draft results line up with the board

I've thought of the news too but I remember seeing some weird things like a few players who were low on the big board get announced first and then it goes to top players back on through to the bottom again.

But yeah it does mostly follow the big board order. Which makes me suspect that you're probably correct but something else weird is happening that we might not understand.

Posted by Benis on 11/24/2020 2:25:00 PM (view original):

Posted by mullycj on 11/24/2020 1:27:00 PM (view original):
Benis - how do you know how much they moved if the big board is never updated? (I have a theory but would like to hear your answer first)

Yeah I'm saying it is updated. Unless I'm a dingus (very possible) and messed it up before..

We will find out for sure in a couple days after Phelan NT is finished.

What's your theory?

its updated after the NT game

Posted by cubcub113 on 11/24/2020 9:17:00 PM (view original):

Posted by Benis on 11/24/2020 2:25:00 PM (view original):

Posted by mullycj on 11/24/2020 1:27:00 PM (view original):
Benis - how do you know how much they moved if the big board is never updated? (I have a theory but would like to hear your answer first)

Yeah I'm saying it is updated. Unless I'm a dingus (very possible) and messed it up before..

We will find out for sure in a couple days after Phelan NT is finished.

What's your theory?

its updated after the NT game

Lamar Terry for OSU jumped from 96 to 88 last year after he won the title, I distinctly remember.

I'll ask Pallas to back me up here.

8.5.2

True.

He made the game-winning basket with only a few seconds left and moved up the Big Board a handful of spots.

2 notes and/or questions.

I have seen players whose status was "Likely Staying" change after the player declared for the draft.

I cant remember the exact number but around #28 a Junior changes from "Likely Going" to "On The Fence" When I think of on the fence I think 40-60% but it sounds like lot of you think that 70+% is still on the fence. What percentage do most of you consider on the fence?