“well let me just say, im curious what your 5 criteria are, but basically being top 5 by position has no impact on the draft big board. obviously, top 5 players are likely to have higher ratings and go - however, there is 0 direct effect there. by considering ratings (assuming you do so with an accurate measure), you've fully accounted for the indirect impact being top 5 by position could have. same goes for overall rating.”

I agree there is no direct relationship as to recruit rank or OVR. I’m just sharing my own personal “rule of thumb” for assessment, some version of which I’ve shared a few times in the past. Because top 5 are so likely to have the ratings that make them early entry candidates, it’s an excellent place to start for assessment, IMO. Not meant to imply the system takes that ranking into account.

There are examples of players who may meet multiple criteria, and still are fairly obviously not EE candidates. But my reasoning is that *among the pool of players high D1 teams are likely to be interested in* those types of players are not common.

From another thread (I have slightly tweaked from the first time I posted this last year, but the essence is the same):

1) Class ranking by position. *Legitimate* Top 5 players by position are very likely to be on the big board prior to their senior year, and are EE candidates. Top 20 players by position are all in the watch zone for me. There are obviously some duds that never get too close, and you can spot them by potential. At the same time, there are (rarely) players out of the top 20 by position with high potential who can get to the big board prior to senior year, with maximized development.
2) LP/Per combo. If either projects to 90+, or if in combination they project to 140+, they may be early entry candidates, if some other baseline conditions are met.
3) Ath/Spd/Def core. In general, a 90 average (270 for pg, sg, and sf, 180 for pf and C) can put the player in range, if other conditions are met.
4) Skill cores average above 80. For guards, it’s per, ball handling, pass; for bigs it’s rebounding, block, and LP. Keep in mind, the fake NBA GMs value skill cores (especially LP and per) more than the HD community does.
5) Overall near 800. This is the least important of the indicators, *notably for players with high work ethic and/or durability*, but especially if one or more of the indicators above is true, this one can be worth looking at, especially if the player has no glaring weakness at his position.

The 3rd above is the shakiest for me, because as you allude to, SF are tougher to evaluate, and the system doesn’t care about the player’s listed position, or how you happened to use him. Physical core ratings are super important to us in how we evaluate and build winning teams, and I think we sometimes project that valuation into the system a little more than it’s actually worth *to the big board/draft*.

Posted by gillispie1 on 6/2/2019 9:39:00 PM (view original):

Posted by Benis on 6/1/2019 2:47:00 PM (view original):
Gil- how would one plan for a Sophomore EE who was never listed on the big board?

Asking for a friend.

well, that is not impossible, but that is like a 1%, probably less. of the hundred + sophs on big 6 teams, i really doubt 2/season go. although, people doing better at draft planning (which should be expected over time), it could be more likely than it used to be (because more players in the top 100 are kept in statistically better ranged). it used to be like 1% on juniors, and significantly less on sophs - doubt that has moved a ton.

short answer, you don't - not directly. its just too unlikely to build around. by in general working on depth, and by valuing multi-position players, you are really doing all you can.

Yeah, the chances are really low (<1%), which is why it SUCKS when you get one.

Gil- I have a pretty good idea of what type of players are "EE worthy" and which aren't. Understanding that will tell you decently where they'll fall on the big board and then chance they'll leave early. I get all that and as you say, it's really not all that hard.

Based upon what I've tracked, I've still seen some funny things come out from EE results.

1. A Jr ranked 40th on the big board leaves early at the same rate as 95th.
2. Out of 100 Sophomores ranked between 50 and 70, only 5% left early. However, out of 77 Sophomores ranked between 84-100, 12% left early.

You could argue that my sample size isn't huge, but it seems big enough to draw some conclusions from. Basically, after the top 30 players, it's a crap shoot. You might as well not bother trying to withhold key ratings or influence anything after that point.

Posted by topdogggbm on 5/31/2019 7:23:00 PM (view original):

Posted by shoe3 on 5/31/2019 6:00:00 PM (view original):
Let’s test your claim. If you get into signing range at high for a player you really want, do you just stop until your rival beats you back to moderate? Or do you do more, to try to get a better shake? If you keep extending effort for a player you’re trying to win, doesn’t that mean you think your individual battle is something other than 50/50?

What you’re talking about is the effects of a small sample size, and limited perspective. It doesn’t make the odds for a 71-29 battle 50/50 (they’re not) but it does affect how a person approaches battling, for sure.

That's a pretty good post. Of course i continue my effort. But i guess what i'm getting at is that the HD odds have not proven to be accurate by any means, in my time playing. And we never have discussions here about how accurate they are for everybody. All discussions are based on "i can't believe i lost this roll again and again and again when i led".

If you tell me that i have an 80% chance to win my next scratch off, i'm stopping in the middle of dinner and i'm sprinting to the store while running every stop sign to buy one. If you tell me that i have an 80% chance to sign a recruit in HD, i tell you "that's nice", while getting two back up players unlocked.

So my "discomfort" with 80% applies to the HD engine. And not real life, or Poker as kcsundevil mentioned.

I'm not really sure what proof you have that HD odds are not accurate. Is it purely based on forum interaction? That's not very reliable, particularly with a small sample size, because people are going to complain (myself included) way more than someone boasting about winning a 80/20 split. This is something you deal with in quality control. People tend to have more energy to write negative reviews than they do positive ones.

What you are describing is your particular mindset on how you handle bad beats. Which is fine. Clearly it is working for you, I don't think anyone is doubting that.

This does apply to the poker though. I can tell you'd be an incredibly tight player ;)

Posted by Basketts on 6/3/2019 12:07:00 PM (view original):

Posted by topdogggbm on 5/31/2019 7:23:00 PM (view original):

Posted by shoe3 on 5/31/2019 6:00:00 PM (view original):
Let’s test your claim. If you get into signing range at high for a player you really want, do you just stop until your rival beats you back to moderate? Or do you do more, to try to get a better shake? If you keep extending effort for a player you’re trying to win, doesn’t that mean you think your individual battle is something other than 50/50?

What you’re talking about is the effects of a small sample size, and limited perspective. It doesn’t make the odds for a 71-29 battle 50/50 (they’re not) but it does affect how a person approaches battling, for sure.

That's a pretty good post. Of course i continue my effort. But i guess what i'm getting at is that the HD odds have not proven to be accurate by any means, in my time playing. And we never have discussions here about how accurate they are for everybody. All discussions are based on "i can't believe i lost this roll again and again and again when i led".

If you tell me that i have an 80% chance to win my next scratch off, i'm stopping in the middle of dinner and i'm sprinting to the store while running every stop sign to buy one. If you tell me that i have an 80% chance to sign a recruit in HD, i tell you "that's nice", while getting two back up players unlocked.

So my "discomfort" with 80% applies to the HD engine. And not real life, or Poker as kcsundevil mentioned.

I'm not really sure what proof you have that HD odds are not accurate. Is it purely based on forum interaction? That's not very reliable, particularly with a small sample size, because people are going to complain (myself included) way more than someone boasting about winning a 80/20 split. This is something you deal with in quality control. People tend to have more energy to write negative reviews than they do positive ones.

What you are describing is your particular mindset on how you handle bad beats. Which is fine. Clearly it is working for you, I don't think anyone is doubting that.

This does apply to the poker though. I can tell you'd be an incredibly tight player ;)

I collected data on battles awhile back. Here are the results.

Posted by Benis on 6/3/2019 8:53:00 AM (view original):
Gil- I have a pretty good idea of what type of players are "EE worthy" and which aren't. Understanding that will tell you decently where they'll fall on the big board and then chance they'll leave early. I get all that and as you say, it's really not all that hard.

Based upon what I've tracked, I've still seen some funny things come out from EE results.

1. A Jr ranked 40th on the big board leaves early at the same rate as 95th.
2. Out of 100 Sophomores ranked between 50 and 70, only 5% left early. However, out of 77 Sophomores ranked between 84-100, 12% left early.

You could argue that my sample size isn't huge, but it seems big enough to draw some conclusions from. Basically, after the top 30 players, it's a crap shoot. You might as well not bother trying to withhold key ratings or influence anything after that point.

You are pretty much on the money. You want to be below 30 on an adjusted basis (NT success) which means if you are a favorite type, better to stay like 40 or so or below. The juniors at 40 vs 80 don’t matter (depending on the year and the RNG, the guys may leave enough at the top that folks in that bottom range - probably more like 90+ - have a shot at not even rolling a die).

However, 20 vs 40 or 15 vs 25 for sophs is a huge difference. Really that is where the planning comes in, you are spot on in your observations, but lots of folks do no draft planning which is a huge hit. And it’s not just about planning within one season, it’s something you want to take back to recruiting, which I think even fewer people do.

The soph thing you mention seems kinda unlikely, but it totally could be right. I can’t say I’ve had nearly enough sophs on the board to have a good sense it’s not true, it might be something you might want to follow up on. It’s not impossible in a game like this that something nonsensical like that could happen. That is quite a few sophs. It is true that a 10% RNG will yield stranger seeming results over 100 players than a 30% or 70%, and if you are cherry picking those sub ranges (as opposed to saying like, what is 50-70 vs 70-100), it’s way more likely you can find something that doesn’t make sense

Sweet. That's good info. Thanks Benis.

Posted by gillispie1 on 6/2/2019 9:36:00 PM (view original):

Posted by topdogggbm on 6/1/2019 4:11:00 PM (view original):

Posted by shoe3 on 6/1/2019 7:48:00 AM (view original):
Benis sure did a number on you, dog. Pity.

He sure did. I've won 4 titles in 6 months and might double that in the next 3 months with my current rosters. Just because i feel different about factors in this game that can't be controlled, doesn't mean i don't know what i'm talking about. Everybody said Steph was too small. He's doing fine. Lonzo don't shoot right, he's doing fine. My opinions are my opinions. I do a lot for this site. I mentor a ton of new guys here. And they're getting successful as well. I don't push my theory on them. Because it's not a fact. It's my opinion. An opinion i was just sharing here. I knew it would fire up the troops. And it did. Any of you that think i am foolish, look me up in the D2 1x worlds. We'll set up a 10+ game home and home series if you like. My open offer is on the table.

just giving you a hard time :) i totally agree that a 70/30 should basically be considered a coin flip, not counted on.

My man! Thank you. This was my overall point. And thanks for being a good sport. It's not that serious. The odds just aren't something that's dependable in this game.

But i do love to stir up the pot here every now and then. I won't lie

Posted by Basketts on 6/3/2019 12:07:00 PM (view original):

Posted by topdogggbm on 5/31/2019 7:23:00 PM (view original):

Posted by shoe3 on 5/31/2019 6:00:00 PM (view original):
Let’s test your claim. If you get into signing range at high for a player you really want, do you just stop until your rival beats you back to moderate? Or do you do more, to try to get a better shake? If you keep extending effort for a player you’re trying to win, doesn’t that mean you think your individual battle is something other than 50/50?

What you’re talking about is the effects of a small sample size, and limited perspective. It doesn’t make the odds for a 71-29 battle 50/50 (they’re not) but it does affect how a person approaches battling, for sure.

That's a pretty good post. Of course i continue my effort. But i guess what i'm getting at is that the HD odds have not proven to be accurate by any means, in my time playing. And we never have discussions here about how accurate they are for everybody. All discussions are based on "i can't believe i lost this roll again and again and again when i led".

If you tell me that i have an 80% chance to win my next scratch off, i'm stopping in the middle of dinner and i'm sprinting to the store while running every stop sign to buy one. If you tell me that i have an 80% chance to sign a recruit in HD, i tell you "that's nice", while getting two back up players unlocked.

So my "discomfort" with 80% applies to the HD engine. And not real life, or Poker as kcsundevil mentioned.

I'm not really sure what proof you have that HD odds are not accurate. Is it purely based on forum interaction? That's not very reliable, particularly with a small sample size, because people are going to complain (myself included) way more than someone boasting about winning a 80/20 split. This is something you deal with in quality control. People tend to have more energy to write negative reviews than they do positive ones.

What you are describing is your particular mindset on how you handle bad beats. Which is fine. Clearly it is working for you, I don't think anyone is doubting that.

This does apply to the poker though. I can tell you'd be an incredibly tight player ;)

I have no proof, no data, no studies, nothing. I wasnt even trying to say that the game is incorrect in its methods (ex... in a thousand 61-39 battles, the leader only wins 50% of the time, creating "wrong" odds).

i was just talking out loud about my opinions. Nothing mathematical or statistical. Just listening to coaches talk. Seeing what i've dealt with. Things like that. I agree that we will hear TONS more people complain when they lose (including myself also) than when they win. But even when there's no battling going on, coaches have discussed "looking back". And saying, "yea i've lost 7 straight while leading". So i just don't put my faith in the odds and i believe everything is a toss up (when it's only 2 teams involved). That way i never get upset when i lose a player when i have s giant lead. And i don't get excited when i win a battle i shouldn't have even been in.

With an 80/20 lead in ANYTHING else in life, i'm comfy. But not HD. I'm now officially done spilling my thoughts on this. (And i don't play poker. I've never had a feel for playing cards)

Posted by shoe3 on 6/3/2019 6:32:00 AM (view original):
“well let me just say, im curious what your 5 criteria are, but basically being top 5 by position has no impact on the draft big board. obviously, top 5 players are likely to have higher ratings and go - however, there is 0 direct effect there. by considering ratings (assuming you do so with an accurate measure), you've fully accounted for the indirect impact being top 5 by position could have. same goes for overall rating.”

I agree there is no direct relationship as to recruit rank or OVR. I’m just sharing my own personal “rule of thumb” for assessment, some version of which I’ve shared a few times in the past. Because top 5 are so likely to have the ratings that make them early entry candidates, it’s an excellent place to start for assessment, IMO. Not meant to imply the system takes that ranking into account.

There are examples of players who may meet multiple criteria, and still are fairly obviously not EE candidates. But my reasoning is that *among the pool of players high D1 teams are likely to be interested in* those types of players are not common.

From another thread (I have slightly tweaked from the first time I posted this last year, but the essence is the same):

1) Class ranking by position. *Legitimate* Top 5 players by position are very likely to be on the big board prior to their senior year, and are EE candidates. Top 20 players by position are all in the watch zone for me. There are obviously some duds that never get too close, and you can spot them by potential. At the same time, there are (rarely) players out of the top 20 by position with high potential who can get to the big board prior to senior year, with maximized development.
2) LP/Per combo. If either projects to 90+, or if in combination they project to 140+, they may be early entry candidates, if some other baseline conditions are met.
3) Ath/Spd/Def core. In general, a 90 average (270 for pg, sg, and sf, 180 for pf and C) can put the player in range, if other conditions are met.
4) Skill cores average above 80. For guards, it’s per, ball handling, pass; for bigs it’s rebounding, block, and LP. Keep in mind, the fake NBA GMs value skill cores (especially LP and per) more than the HD community does.
5) Overall near 800. This is the least important of the indicators, *notably for players with high work ethic and/or durability*, but especially if one or more of the indicators above is true, this one can be worth looking at, especially if the player has no glaring weakness at his position.

The 3rd above is the shakiest for me, because as you allude to, SF are tougher to evaluate, and the system doesn’t care about the player’s listed position, or how you happened to use him. Physical core ratings are super important to us in how we evaluate and build winning teams, and I think we sometimes project that valuation into the system a little more than it’s actually worth *to the big board/draft*.

i think the reality of games like this is the model that gets you close to the answer is the one that is best for you. there are times i talk to folks and its like this is a known quantity, we know what the model or structure of the formula, or whatever, is exactly - so go calculate XYZ and act accordingly. other times its like well, i think of it like this, but i have no idea if it actually works that way under the hood - i just know that it gets me to a place where i can predict results well, so i go with it. whether the model is 'real' or not, if it gets you from A to B, who cares?

i think your example is a model of the latter. i wouldn't recommend other folks necessarily pick that up, because we do know its a ratings based thing and its not that hard to get a feel for how ratings map to draft board rankings - once you know that is the underlying model and that is what you should be doing, in theory. but if you already have a model that works for you, i don't really see a need to change.

that said, i would recommend other folks try to work off ratings, because generally speaking, a model that is more closely aligned with the thing its modelling, its going to be more accurate, so when we have the luxury of knowing how something really is, we may as well try to build understanding around that. sometimes its worth using a less-aligned model for simplicity, but i don't think that is the case here.

your insight on #4 is important - lp/per are way overvalued in the big board relative to how people value them (in many cases), along with the various skill ratings for players who don't really need them. like passing on pg? way more valuable than it reflects on the big board. passing on a SF through C? way more valuable on the big board than in reality. understanding those deviation is definitely key to avoiding too many draft picks - or at least, the key to avoiding 'good' guys leaving early, which is the killer, its one thing to lose the elite ones early.

so i think overall what you are doing fundamentally is right, which is saying hey, i dont really need my <whatever> to have <XYZ> because it adds little value in reality while adding a lot to big board ranking. so, screw that. that is a principle conclusion for both of us, regardless of how we get there!

" if you are cherry picking those sub ranges (as opposed to saying like, what is 50-70 vs 70-100), it’s way more likely you can find something that doesn’t make sense"

Cherry picking! How dare you. I'd never manipulate data to support my narrative on these forums.

Here are the total results. Ranges were kind of based upon the cut off between Likely going, On the fence, etc. They get a little wonky but I didn't feel like going back and fixing them but I don't think it affects the results or conclusions you would draw from the data. Enjoy everyone.

EE Results

Whew, Steiner math moment.