Quote: Originally posted by vandydave on 11/17/2009treating HD as a math problem =

Come on, that's what makes HD fun.

Quote: Originally Posted By treyomo on 11/17/2009

You also need to factor in the O/D you're playing. If you use a slowdown triangle/zone, athleticism, speed, and stamina don't mean as much as if you are playing uptempo press. I do a similar analysis for GD, and here's how I break it down:

- Decide which 3-4 rankings are the "core" rankings for each position

- Assign those weights depending on importance. that sum up to roughly 80%

- Take the other rankings (excluding WE and durability) and assign those lower weights to reach 20%

- For HD, I'd add one modification by adding pace and O/D set adjustments, reducing the importance of stamina if playing slowdown and increasing if uptempo, and increasing the importance of ath/speed if playing FB/FCP and decreasing if going zone / triangle (which I perceive, right or wrong, as having the least amount of movement and therefore requiring slightly less athleticism)

- When reviewing the list, set minimums for each core rating that you won't accept below that score

edit - missed the bolded part above, he already identified m first issue with a decent work around

This seems like some pretty good advice regarding how to do such a project. much like it seems rails has, i have written formulas, 2 of the issues involved I see:

#1 - there probably are some intuitive minimums, that could be written into if - then type excel statements, but unless you are a programmer, might be more than you would want to bite off - after this realization is when I quit using forumulas

in d1, as 1 example, many top coaches would not recruit a PG with 44 passing , even if he was a 750 player, with 90 ath / sp / per / bh / def - but certainly some formula might grade him out as near the best player in the country??? emphasis on the word might - although I just recently saw a similiar recruit and was mighty tempted

#2 - I think b4 potential was introduced, some coach was coming pretty close to writing code that was predicting how a player would turn out, but with the hi-med-lo improvement rates, these formulas are at least of less value, unless those FFS rates are also included, again, can be overcome with programming, to an extent at least

Quote: Originally Posted By oldresorter on 11/17/2009

Quote: Originally Posted By treyomo on 11/17/2009

You also need to factor in the O/D you're playing. If you use a slowdown triangle/zone, athleticism, speed, and stamina don't mean as much as if you are playing uptempo press. I do a similar analysis for GD, and here's how I break it down:

- Decide which 3-4 rankings are the "core" rankings for each position

- Assign those weights depending on importance. that sum up to roughly 80%

- Take the other rankings (excluding WE and durability) and assign those lower weights to reach 20%

- For HD, I'd add one modification by adding pace and O/D set adjustments, reducing the importance of stamina if playing slowdown and increasing if uptempo, and increasing the importance of ath/speed if playing FB/FCP and decreasing if going zone / triangle (which I perceive, right or wrong, as having the least amount of movement and therefore requiring slightly less athleticism)

- When reviewing the list, set minimums for each core rating that you won't accept below that score

edit - missed the bolded part above, he already identified m first issue with a decent work around

This seems like some pretty good advice regarding how to do such a project. much like it seems rails has, i have written formulas, 2 of the issues involved I see:

#1 - there probably are some intuitive minimums, that could be written into if - then type excel statements, but unless you are a programmer, might be more than you would want to bite off - after this realization is when I quit using forumulas

in d1, as 1 example, many top coaches would not recruit a PG with 44 passing , even if he was a 750 player, with 90 ath / sp / per / bh / def - but certainly some formula might grade him out as near the best player in the country??? emphasis on the word might - although I just recently saw a similiar recruit and was mighty tempted

#2 - I think b4 potential was introduced, some coach was coming pretty close to writing code that was predicting how a player would turn out, but with the hi-med-lo improvement rates, these formulas are at least of less value, unless those FFS rates are also included, again, can be overcome with programming, to an extent at least

I think this is where the eyeball test has to come in. I wouldn't want a formula that would automatically reject this kid based on the 44 passing. Sure, I'd probably never play him at PG, but I would absolutely recruit him and stick him at SG (or SF depending on his REB).

This post could not be converted. To view the original post's thread, click here.

I developed my own crazy formula a while ago. But my opinion is that I think it is just about worthless* in recruiting these days. Before potential I thought it was very helpful. But now I pretty much try to find players that have high potential ratings in as many key ratings as possible. I don't even pay all that much attention to the starting numbers, primarily because if a player has high potential (especially in D3) the player is almost always going to have an acceptable starting baseline.

* I say worthless -- but there could still be some value to a formula if you tried to make some potential projections with it. But I'm not about to take the time to recalibrate every green and red colored cell by hand and I don't know how to automate it, if that's even possible.

I still do use my formula to help me figure out starting lineups and depth charts. But as a recruiting tool, I've found it really doesn't help that much any more.

Quote: Originally Posted By kujayhawk on 11/17/2009I developed my own crazy formula a while ago. But my opinion is that I think it is just about worthless* in recruiting these days. Before potential I thought it was very helpful. But now I pretty much try to find players that have high potential ratings in as many key ratings as possible. I don't even pay all that much attention to the starting numbers, primarily because if a player has high potential (especially in D3) the player is almost always going to have an acceptable starting baseline.

* I say worthless -- but there could still be some value to a formula if you tried to make some potential projections with it. But I'm not about to take the time to recalibrate every green and red colored cell by hand and I don't know how to automate it, if that's even possible.

I still do use my formula to help me figure out starting lineups and depth charts. But as a recruiting tool, I've found it really doesn't help that much any more

I made a post, then saw KU's post. I'm in the same boat. Excel went out the window when potential arrived.

To the OP, I'll take a look at the criteria you're looking for and get back to you this afternoon or tomorrow morning.

I use some rule of thumbs, off and on when recruiting, somewhat like what you have, but weighted somewhat,

as an obvious (i think obvious at least), REB is weighted higher in true rebounding than say SP, and this is not the formula I use, but I use something like true REB = (3*REB + .5*SP + 1.5*ATH) / 5 - I could justify the use or non use of several other variables in this one simple calcuation with a near infinite number of weightings

in these attempts we all are guessing, guessing how tarek wrote the code, and as far as I know, unless some math genius used some real sophistacated multivariable regression analysis, we won't figure it out.

HOnestly, I would usually be using the thing after I had done my first eyeball test. I just like having multiple ways to look at several players and figure tyhings out. . .



And I do do potential projectections(As ranges, low/high/average) as well.

(If they get about as little as possible improvement in all categories, if they get about as high as can be expected given potential in each categoriy, and the average of the high and low. Trying to only draft a player if I think they coudl at least contribute off the bench if they get the absolute worst their potential allows for - main cutoff.

3 random comments, in which I'll disagree slightly with two coaches who are a lot better than me.

1) I would disagree with Rails about blks not being an important part of LP defense. I would say it might be the most important component, especially if you are running zone.

2) I would add to OR's rebounding formula- work ethic. I know that admins have said for years work ethic doesn't help on the court, but it can't be. All other things close to equal the high WE guy will out rebound the low WE guy every time, and by monster margins. It seems every time I've had a super high WE postman he's become a beast on the boards.

3) Sadly, I'm not sure these formulas would help you, in fact, I think they will hurt you. Too many new coaches I think are worried about getting "the best players". Formulas are there for when you can afford to split hairs, as in who was better Babe Ruth or Ted Williams. In recruiting, especially when starting out you really should stay within 100 miles, 170-180 at the very most. And in that are there are only going to be so many good players, it won't be hair splitting deciding who is best. Just get the very best players in your area, regardless of everything else, run'em out on the court and hope for the best. That's always been my theory.

But of course Rails and OR have about 25-30 more national titles than I do so you might want to ignore the above.

Quote: Originally Posted By dalter on 11/17/2009Fair enough to be called out on that. As VD alluded to, it just offends my sensibilities.

Agreed, I have looked at many things others have done regarding numbers and those guys have done a great job with some of the things they have done, but in the end I am fairly confident my judgement and gut is going to be a lot better then what a spreadsheet can tell me.

Quote: Originally Posted By arssanguinus on 11/17/2009PArt of what I am thinking is that sometimes when I am looking for a player I am looking to fill a very specific role - IE: THis team is really weak at rebounding, so I need another rebounder

This is key to recruiting imo, which makes formulas difficult.

Quote: Originally Posted By daveymac on 11/17/2009
3 random comments, in which I'll disagree slightly with two coaches who are a lot better than me.

1) I would disagree with Rails about blks not being an important part of LP defense. I would say it might be the most important component, especially if you are running zone.

2) I would add to OR's rebounding formula- work ethic. I know that admins have said for years work ethic doesn't help on the court, but it can't be. All other things close to equal the high WE guy will out rebound the low WE guy every time, and by monster margins. It seems every time I've had a super high WE postman he's become a beast on the boards.

3) Sadly, I'm not sure these formulas would help you, in fact, I think they will hurt you. Too many new coaches I think are worried about getting "the best players". Formulas are there for when you can afford to split hairs, as in who was better Babe Ruth or Ted Williams. In recruiting, especially when starting out you really should stay within 100 miles, 170-180 at the very most. And in that are there are only going to be so many good players, it won't be hair splitting deciding who is best. Just get the very best players in your area, regardless of everything else, run'em out on the court and hope for the best. That's always been my theory.

But of course Rails and OR have about 25-30 more national titles than I do so you might want to ignore the above.

I have seen the same thing with WE, if you notice, I said I probably could add a few other ratings to the calculation and that my example was not how I did things, I also mentioned that we are trying to guess what tarek did, we really aren't going to know unless seble, who is more open, comes out and tells us which my guess is some of us would love, others of us would find not so good - to me, it is fun trying to figure out

I think a formula becomes more usable the higher up the food-chain you go in HD. At DIII I really think you can nearly throw out formulas due to the simple fact that every player will have a glaring weakness or two, and thus the only ones who will qualify may be the average guy.

At DIII I want the guy who can score, the guy who can play D, the guy who can rebound, and the guy who can hold on th the friggin ball and NOT turn it over. But I can't get all of that in one guy... I have to have some of each on my team and hope they mesh to form a competetive team.

*edit... on the WE issue. several years ago I studied RB in alot of detail across several worlds,,, and I came away convinced there was a direct connect between WE and RB... now, where the connection is may be another story. Cause or effect... LOL, one of the lil secrets to keep us guessing.

I also like drafting small skill set chunks at dIII. . .

you gotta pick your warts