Draft Rater

nydownunder · Post by **nydownunder** » Thu Oct 05, 2006 8:22 am

Ok guys. I'm bored and curious so I am going to endeavor on building a spreadsheet to rate some or everyone's drafting ability.

Before doing so, I want some opinions regarding some thoughts of mine. Two of the inputs will obviously be (my) projected stats and the actual stats of the players converted into a value system which matches up pretty well with actual standing values (ie 115 points).

So any player performaing over and above his stats is a POSITIVE. However, shouldn't there be some value in drafting a player that performs slightly less than projections, yet was drafted 2 rounds later than the avg of 22 leagues? Also, should it be a POSITIVE if that player underperforms projections, but outperforms the average score of players drafted in the some spot? Also, is the greatest value (in drafting) when you drafted a player earlier than the avg of 22 leagues and in which he outperformed his projections?

I'm thinking perhaps I break the value (Actual-Proj) of a particular players draft into 2 or 3 parts: Allocation, Selection, Interaction. Allocation could be relative to other players drafted in the same spot; Selection having to do with whether you drafted him earlier or later relative to the league averages; and interaction being the difference in added value (Act-Proj) less Allocation and Selection.

WOW! I think I jsut bit off more than I can handle. Keep in mind that without knowing how much each manager actually played these players, this is nowhere near an exact science.

Fee free to add comments and give me about a year to compile.

Spyhunter · Post by **Spyhunter** » Thu Oct 05, 2006 9:24 am

Why not use the scores from ESPN's player rater? it is a pretty good system that normalizes performance. Developing such as system is actually pretty hard - for example see how STATS screws it up by rating someone who has a ba of 1.000 (1 for 1) above someone who hits .340 ....

nydownunder · Post by **nydownunder** » Thu Oct 05, 2006 9:44 am

Originally posted by Spyhunter:

Why not use the scores from ESPN's player rater? it is a pretty good system that normalizes performance. Developing such as system is actually pretty hard - for example see how STATS screws it up by rating someone who has a ba of 1.000 (1 for 1) above someone who hits .340 .... Because it doesn't provide value for drafting someone ahead of the 22 keague average (ie knowing someone would outperorm. The theory being those who grabbed a guy early (and outperformed) knew he would outperform and those whom got a player 2 rounds after the 22 league average (who outperformed) had more to do with luck (ie if he really rated that player, he would have grabbed him earlier than the 22 league average.

It eesentially will reward those managers whom went after those outperfomers rather than those who had them fall to them. Yeah, I understand guys falling to someone can be deemed skill, but on avg I would think it had more to do with luck of the draft.

CC's Desperados · Post by **CC's Desperados** » Thu Oct 05, 2006 3:40 pm

Originally posted by nydownunder:

quote:Originally posted by Spyhunter:

Why not use the scores from ESPN's player rater? it is a pretty good system that normalizes performance. Developing such as system is actually pretty hard - for example see how STATS screws it up by rating someone who has a ba of 1.000 (1 for 1) above someone who hits .340 .... Because it doesn't provide value for drafting someone ahead of the 22 keague average (ie knowing someone would outperorm. The theory being those who grabbed a guy early (and outperformed) knew he would outperform and those whom got a player 2 rounds after the 22 league average (who outperformed) had more to do with luck (ie if he really rated that player, he would have grabbed him earlier than the 22 league average.

It eesentially will reward those managers whom went after those outperfomers rather than those who had them fall to them. Yeah, I understand guys falling to someone can be deemed skill, but on avg I would think it had more to do with luck of the draft. [/QUOTE]Mark, The best result for this will be next years draft. There are some players owners will reach for. If I took Hanley Ramirez in the 23 rd, another owner takes him in the 14 round. You want to reward the more aggressive owner. No matter where we took him, we both ended up with 2nd round talent. One owner might have needed his 2nd middle infielder, the other his 3rd. There are certain players each year who will make an impact. The key is finding as many as possible. If you project stats, you would be lucky to be right 50 % of the time. If you talk player performance from year to year, the a difference story. The difference from winning and losing could be only two or three players. If you review your own draft, the best first three picks would have been Reyes, Soriano, and Berkman to build your team. As Cabrara and Rollins didn't hurt you, Pedro did. When I look at my own draft, Vlad, Berkman, and Giles....if I just replaced Giles stats with Sorianos, you could see the gain across the board. That one player might have given me 200 more hitting points. I didn't go to the draft wanting him, but I'm not sure if I would have passed on him in round two. I did want Berkman. I believe Vinnie's team had a perfect first three picks. Reyes instead of Crawford would have been better, but that is easier to say now. You can review draft result as much as you want, but I don't think you are going to find the keys to success there. The player pool changes so much from year to year. I think this year, many of us thought the catching pool was short on talent. Next year, It might look a lot better. This year saves were down and steals were up. Starting pitching was down across the board. Good luck with your research, call me or send PM if you want more details on this.

[ October 05, 2006, 09:50 PM: Message edited by: CC's Desperados ]

bjoak · Post by **bjoak** » Thu Oct 05, 2006 10:15 pm

Good idea and you are the guy to do it. My suggestion is to not worry about where a player is drafted in relation to where others drafted him. Value is value. To me if someone drafted Thomas in the fifth round and that demonstrates value compared to other fifth rounders that is credit. In other words, someone should not lose credit for picking a player earlier than others if he was worth more than other players of the same round. By the same token someone should lose credit if they drafted, oh, Eric Chavez later than everyone else but he still was a loss at that round. If you're evaluating players properly you should not lose credit because you are picking players at a more appropriate spot than conventional wisdom would suggest.

I look forward to seeing your results.

nydownunder · Post by **nydownunder** » Fri Oct 06, 2006 1:37 am

I hear you bjoak, but there is a difference in someone going after someone (as compared to the other 21 leagues avg) and someone falling to you (that noone wanted in those spots).

I haven't figured it all out yet. I know one thing I don't want to do with this is simply take the values (which arguably the website can provide) and sum them for each manager. If I simply did that, a certain someone from last year would have looked like a genious...but I know he had all the luck in the world IMHO.

The other thing to keep in mind is that this is only approx. 2/3's of the game. In some cases, like mine, I'd say FAAB added 50% of my value for the year. that sort of stuff is too labor and data intensive to sort out.

bjoak · Post by **bjoak** » Fri Oct 06, 2006 5:42 am

Yes, I wouldn't count faab. I thought the purpose was to find out how well people drafted. If you take every other factor into account you may as well just look at the final standings.

nydownunder · Post by **nydownunder** » Fri Oct 06, 2006 6:01 am

I am interested in somehow showing the skill level of drafting. Honestly, I don't know if that is entirely possible but worth a crack. But in order to do so, I need to make assumptions on what constitutes a skilled pick and a lucky pick. I think if someone rated/projected a player higher than the average, then they more than likely drafted them earlier than the average...and vice versa.

It's kind of like investments and comparing to a benchmark. If the benchmark says Oswalt at #60 and you pick him at #54 then arguable your are overweight Oswalt. So an overweight with a positive resul euals a positive. On the flip side, if Eric Chavez is benchmarked at #80 and I pick him at #90 I am underweight Chavez, thus if there is a negative result (which there was), then it would be a positive because you drafted him later relative to others (ie you were smart to not pick him earlier = you weighed the risk reward trade off)....that I think is the best means without sitting down with every person to TRY and judge their skill level.

It's not perfect, just perhaps interesting spin on things.

nydownunder · Post by **nydownunder** » Fri Oct 06, 2006 9:22 am

Originally posted by CC's Desperados:

quote:Originally posted by nydownunder:

quote:Originally posted by Spyhunter:

Why not use the scores from ESPN's player rater? it is a pretty good system that normalizes performance. Developing such as system is actually pretty hard - for example see how STATS screws it up by rating someone who has a ba of 1.000 (1 for 1) above someone who hits .340 .... Because it doesn't provide value for drafting someone ahead of the 22 keague average (ie knowing someone would outperorm. The theory being those who grabbed a guy early (and outperformed) knew he would outperform and those whom got a player 2 rounds after the 22 league average (who outperformed) had more to do with luck (ie if he really rated that player, he would have grabbed him earlier than the 22 league average.

It eesentially will reward those managers whom went after those outperfomers rather than those who had them fall to them. Yeah, I understand guys falling to someone can be deemed skill, but on avg I would think it had more to do with luck of the draft. [/QUOTE]Mark, The best result for this will be next years draft. There are some players owners will reach for. If I took Hanley Ramirez in the 23 rd, another owner takes him in the 14 round. You want to reward the more aggressive owner. No matter where we took him, we both ended up with 2nd round talent. One owner might have needed his 2nd middle infielder, the other his 3rd. There are certain players each year who will make an impact. The key is finding as many as possible. If you project stats, you would be lucky to be right 50 % of the time. If you talk player performance from year to year, the a difference story. The difference from winning and losing could be only two or three players. If you review your own draft, the best first three picks would have been Reyes, Soriano, and Berkman to build your team. As Cabrara and Rollins didn't hurt you, Pedro did. When I look at my own draft, Vlad, Berkman, and Giles....if I just replaced Giles stats with Sorianos, you could see the gain across the board. That one player might have given me 200 more hitting points. I didn't go to the draft wanting him, but I'm not sure if I would have passed on him in round two. I did want Berkman. I believe Vinnie's team had a perfect first three picks. Reyes instead of Crawford would have been better, but that is easier to say now. You can review draft result as much as you want, but I don't think you are going to find the keys to success there. The player pool changes so much from year to year. I think this year, many of us thought the catching pool was short on talent. Next year, It might look a lot better. This year saves were down and steals were up. Starting pitching was down across the board. Good luck with your research, call me or send PM if you want more details on this. [/QUOTE]Point taken but I do think players who are reached for are reched for for a reason. In the end it is extremely difficult to make such an evaution, just thought I would give it a stab with some assumptions.

But the more I look at it, the more sophisticated it will need to be to provide anything of value. And even it if does provide any value, what's it matter when the Standings really tell the story.

nydownunder · Post by **nydownunder** » Fri Oct 06, 2006 9:38 am

Originally posted by CC's Desperados:

quote:Originally posted by nydownunder:

quote:Originally posted by Spyhunter:

Why not use the scores from ESPN's player rater? it is a pretty good system that normalizes performance. Developing such as system is actually pretty hard - for example see how STATS screws it up by rating someone who has a ba of 1.000 (1 for 1) above someone who hits .340 .... Because it doesn't provide value for drafting someone ahead of the 22 keague average (ie knowing someone would outperorm. The theory being those who grabbed a guy early (and outperformed) knew he would outperform and those whom got a player 2 rounds after the 22 league average (who outperformed) had more to do with luck (ie if he really rated that player, he would have grabbed him earlier than the 22 league average.

It eesentially will reward those managers whom went after those outperfomers rather than those who had them fall to them. Yeah, I understand guys falling to someone can be deemed skill, but on avg I would think it had more to do with luck of the draft. [/QUOTE]Mark, The best result for this will be next years draft. There are some players owners will reach for. If I took Hanley Ramirez in the 23 rd, another owner takes him in the 14 round. You want to reward the more aggressive owner. No matter where we took him, we both ended up with 2nd round talent. One owner might have needed his 2nd middle infielder, the other his 3rd. There are certain players each year who will make an impact. The key is finding as many as possible. If you project stats, you would be lucky to be right 50 % of the time. If you talk player performance from year to year, the a difference story. The difference from winning and losing could be only two or three players. If you review your own draft, the best first three picks would have been Reyes, Soriano, and Berkman to build your team. As Cabrara and Rollins didn't hurt you, Pedro did. When I look at my own draft, Vlad, Berkman, and Giles....if I just replaced Giles stats with Sorianos, you could see the gain across the board. That one player might have given me 200 more hitting points. I didn't go to the draft wanting him, but I'm not sure if I would have passed on him in round two. I did want Berkman. I believe Vinnie's team had a perfect first three picks. Reyes instead of Crawford would have been better, but that is easier to say now. You can review draft result as much as you want, but I don't think you are going to find the keys to success there. The player pool changes so much from year to year. I think this year, many of us thought the catching pool was short on talent. Next year, It might look a lot better. This year saves were down and steals were up. Starting pitching was down across the board. Good luck with your research, call me or send PM if you want more details on this. [/QUOTE]Point taken but I do think players who are reached for are reched for for a reason. In the end it is extremely difficult to make such an evaution, just thought I would give it a stab with some assumptions.

But the more I look at it, the more sophisticated it will need to be to provide anything of value. And even it if does provide any value, what's it matter when the Standings really tell the story.