tysok
All Star team
November 01, 2002 at 09:49PM View BBCode
I know we've went through this before. :)
Since this is the first season that the awards have actually happened I'm going to look at them and point out some "wrongs".
First off I'm dealing with hitters right now, starters and backups for every position in AL and NL.
In the AL I agree with 6 of the 8 starters, 1 of the remaining 2 may be right, I can't decide.
I can see where the other 1 comes from, which is the actual position played instead of just OF. It named Shano Collins as the starter, and Willie Hummel as a backup. The reson is under the current "rules" the AL needed a starting RF. The argument against this is simple. OFs play the OF. The real game is voted on OFs, not LF RF and CF. Purely from the numbers side the "All star page" agrees that Hummel SHOULD have been the starter.
The other starter in question is Berdie Tebbets vs Joe Rose. By the current score Tebbets is the winner by 36 points. However, both of these C are remarkably similar. Roughly the same at bats and the same speed. Tebbets knocked 4 more HRs, 4 more 2Bs, had a 3B, scored 18 more times, and drove in 40 more runs... but Rose hit for better average, had 8 more walks and 30 less Ks... Rose had a higher OPS (1.05 to 1.019) higher OBP (.430 to .397) same SLG (.621 to .622) and hit higher (.356 to .325).
Also by the original Runs Created formula Rose created 10 more runs. Base Performance Value Rose was 17 pnts better, and in my Secondary Fantasy pnts (which doesn't take errors or speed into account) he scores 40 pnts higher than Tebbets. By Win Shares they were both the same offensively(26 for Tebbets 25 for Rose). By the "Dollar Value" equation I use Rose is worth almost $1 mroe than Tebbets.
Tebbets should probably be the starter, but it's close. Rose had 10 errors, Tebbets 0.
For the backups there are 3 in question. Al Zarilla, Mongo Flath, Sean Rhodes. I think I know where this one comes from too, it seems to be looking for backups for specific positions, which in my opinion doesn't matter.
Who cares what positions that backups are? They don't play a game, they don't need to play a game, and the whole thing (for a computer simulation) is to recognize the best players, not the best 2 from every postion, the best in each position and the best of the rest.
A simple case is for Blair Blutarski and Hippo Adams, which even the fantasy pnts we use now tell us they're better than any of those three.
One that's a little tougher is knocking Al Zarilla (441.5 pnts) out and putting Danny Ferrer (418.5 pnts) in.
Ferrer hit 25 more HRs, scored 30 less Runs, 4 less 2B, 7 less 3B, but had 30 more RBIs. The K to BB ratio is about even on Zarilla, while Ferrer walked 37 times and struck out 80+. Zarilla had 15 more SBs, but also had 9 more errors. The big difference is in the other things though. Ferrer had higher OPS (.977 to .784) higher OBP (.386 to .380) higher SLG (.610 to .404) higher AVG (.329 to .288) roughly 10 more Runs Created, higher Base Performance Value (85 to 45) more offensive Win Shares (19 to 15) and was valued as worth $5 more.
Ferrer didn't play as much (460 ABs to 620) but he played better. He may not have been more valuable to his team, but he was more of a stand out in general. And the All Stars is about Stand outs, not MVPs.
Now how did I get to these numbers? I did something different thna I was doing. Not agreeing that fantasy was the way to do it (because of it's misleading nature) I went looking. The beauty of this, and what I was looking for, was to be able to take the top 7 players leftover (after taking out the best per position) and put them on the team... no evaluation necessary.
Here's what I came up with.
The Dollar Value Formula I mentioned above is something I found on the internet, someone had done a lot of work to put a formula together that more or less mirrored how the Rotisserie value system worked, I forgot to get his name off the sight and can't find it again. :(
Dollar Value formula is: (0.207304*SB*1.5)+(0.1251003*HR*1.5)+(0.0300868*RBI]*1.5)+(0.035351976*R*1.5)+((((1498.8+H)/(5450+AB))-0.275)/0.002*3.14*0.8*1.5)
The 1.5 is obviously a place filler to bring the whole thing up to alignment... we don't want to see Bonds priced at $4.5. :)
Base Performance Value (if anyone doesn't know) is:
((BB/SO)*20)+(((H/AB)-0.3)/0.003)+((((2B*0.8)+(3B*0.8)+(HR*1.4))/AB)*365*1.25)
Runs Created (again for those that don't know) is:
((H+BB)*(H+2B+(3B*2)+(HR*3)))/(AB+BB)
The fantasy point system (what we're currently using except it seems to be slightly modified from the version I have here) is:
R+H+2B+(3B*2)+(HR*3)+RBI+BB+(SB*2)-SO-E-CS
or R+TB+RBI+BB+(SB*2)-SO-E-CS
Figure those 4 up for every batter, add them together and you get the numbers I've used. In fact it was just changed to include the last "Fantasy score" in the equation... it was using another formula of mine but I just found it better to do this way, all numbers basically were the same except it did answer the Birdie Tebbets Joe Rose problem, Tebbets was losing by 3 pnts before i changed it, now he is winning by 9 pnts.
Bob
November 01, 2002 at 10:13PM View BBCode
Interesting stuff, Tysok. I actually think your post points to the effectiveness of the current selection process. Putting aside the positional thing wherein there must be a starting RF, LF, and CF (not just 3 OF), your in-depth look came out remarkably similar to Tyson's quick calculations. The only differences are in the very close cases -- and in my opinion that's fine. That's where there should be controversy.
On a similar note, I would love to see some additional statistics kept. Since I'm a Bill James fan, I'd love to see Runs Created and Win Shares. Runs Created would be pretty easy and although a comprehensive Win Shares calculation would be tough, a basis one ought to be doable. I hope that Tyson will add it to his development list.
tysok
November 01, 2002 at 10:20PM View BBCode
This one will be shorter, I think I made my case in the AL. But I do want to point out a few things in the NL...
Rennie Stennett was chosen to be the starter over Gilbert Robinson at 2B. Stennett vs Robinson:
Stennet played more than Robinson this year (687 ABs to 451) scored 18 more runs, 10 more doubles, 5 more triples, stole 37 more bases, but had 11 less errors. Robinson had 18 more HRs and 20 more RBIs, as well as higher OPS (1.001 to .728), higher OBP (.402 to .335) higher SLG (.599 to .393) higher AVG (.322 to .261) 20 more Runs Created, 65 more pnts in Base Performance Value, 10 more pnts in Secondary Fantasy, had more offensive Win Shares (16 to 14) and rated $2 higher.
Under the Fantasy system alone Stennett scores 24 pnts higher. In the combination of formulas I've used Robinson has 60 more points. The only reason Stennet scores higher is his speed, 53 SBs to 16. Speed should be a factor, but not that much.
One other player is Dixie Howell. I can make a case for Stennett to be on the team (in fact he should be) but huh????
Howell hit .253, the lowest on the team. He scored 49 runs, the lowest. He had 21 2B, 21 HRs, 78 RBIs, 31 BB, 92 SO. He had an OBP of .298 again the lowest. He had 284 fantasy pnts, lowest by 90 pnts. Only 65 Runs Created, again the lowest. He had a Base Performance value of 32 pnts, lowest by 12 pnts to Robert Mitchell who should also not be on the team. He had 264 secondary fantasy pnts, lowest by 100 to Johnny Furlong who also doesn't belong. He only had 10 offensive Win Shares, the lowest on the group that should be on the team is Stennett with 14 then it goes to 16. By the Dollar Value he's worth $6.6... lowest by $6 to Dick Stuart who also doesn't belong. By the 4 pnt formula used to get the team he scores 388, more than 200 pnts lower than anyone who SHOULD be on the team. The only exception is Joe Fisk who scored 428 but led all catchers...
I guess it was a bad year for Catchers in the NL. :)
tysok
November 01, 2002 at 10:40PM View BBCode
In a way your right Bob. Controversial yes, wrong no.
In the case of Stennett and Robinson I guess it could be looked at in both ways. The real problems come in though when we force there to be a backup for each position. Dixie Howell is not an all star. When you look at the great OFs we had in the NL like Alan Simmons, Mule Creek, Bobby Thomson, Cal Barnes, Jerry Fernandez, or even Robert Mitchell and Rodney Hamilton... those guys decimate Dixie Howell.
Also on another note. One reason I've worked this hard (don't pitty) on this is I hope it will actually matter later on. The HOF Monitor takes into account the awards a player gets. On some things, like the MVP I don't entirely agree with (the order isn't right but still the right top 3) it can be dealt with. But if the stars, the real stars of Sim Dynasty Willie Mays League are not selected to the All Star game then it throws off the balance of the All Star Award factors in the HOF Monitor as it pertains to SimDynasty.
Couple that with the fact that this IS a game, and we play it in order to finish 1st and get the World Series, or to watch the team grow up, or build it from scratch etc., and that there are no "fans"... the All Star Team becomes another tool like the top10. A way of pointing out the best of the players, and in so doing hopefully rewarding the best of the players with the pnt totals that may get them into the SimDynasty Willie Mays League Hall of Fame.
I love baseball, and I'm a statistician... so this game really appeals to me. On top of that is I don't want my SS to be the 2nd best SS in the league. If there's an OF that did a lot better than my SS but is only 4th in the league I don't want my SS to be in the AS team. I want to see my SS be the best overall player... I want Tim Raines. :) It just doesn't do any good to show that Dixie Howell was the 2nd best SS in the NL when he sucked, and there are 4+ better players out there that just didn't catch. Also 1952 in the NL WML was the exception, usually there may be 2 OFs, this time there were 2 2B, 1 3B, and 4 OFs.
Bob
November 01, 2002 at 11:10PM View formatted
You are viewing the raw post code; this allows you to copy a message with BBCode formatting intact.
I think the actual rankings, while not perfect, are damn good as they are. I agree with you that the issue is who should be on the All-Star team.
To summarize what you're saying:
Currently the starters are 1 each of SP, C, 1B, 2B, 3B, SS, LF, CF and RF. The only change we need there is to select the three best OFs overall and not insist on 1 LF, 1 CF and 1 RF.
Currently for the reserves there are 4 SP and 5 RP -- this is fine in my opinion. For position players there are 1 each of C, 1B, 2B, SS and 3B and 2 OFs (doesn't matter if they're LF, CF or RF). I agree with Tysok that the position players should change to simply the 7 best players left, regardless of position.
geoffrey13
It's kinda funny
November 02, 2002 at 03:20AM View BBCode
The amount of time and energy spent on stats for a league of FANTASY baseball players....total fantasy...nothing even remotely real about them! Yikes!:o
I like it just the way it is...no way should some slugging slug of a RF or LF take the spot of a deserving CF player...CF is an extremely important position defensively.
In the case of my 2B, Gil Robinson, I don't even think he should be there because he was hurt for so long this season...it certainly proved his value to my team because without him we struggled...but the other guy should get the spot.
I do think the voting might take away a bit too much for errors...my 1B Jackie Foxx was IMHO the best offensive player in either league but his 33 errors dropped his votes down....and that's another thing....he has a good arm and poor range...a 1B with poor range simply can't get to a ball so how can he get errors for it? He doesn't make throwing errors because of his good arm so what then is he doing, dropping everything? There should be a glove rating as well in that case, because otherwise it makes no sense to me how he can have so many errors.
geoffrey13
It's kinda funny
November 02, 2002 at 03:20AM View BBCode
The amount of time and energy spent on stats for a league of FANTASY baseball players....total fantasy...nothing even remotely real about them! Yikes!:o
I like it just the way it is...no way should some slugging slug of a RF or LF take the spot of a deserving CF player...CF is an extremely important position defensively.
In the case of my 2B, Gil Robinson, I don't even think he should be there because he was hurt for so long this season...it certainly proved his value to my team because without him we struggled...but the other guy should get the spot.
I do think the voting might take away a bit too much for errors...my 1B Jackie Foxx was IMHO the best offensive player in either league but his 33 errors dropped his votes down....and that's another thing....he has a good arm and poor range...a 1B with poor range simply can't get to a ball so how can he get errors for it? He doesn't make throwing errors because of his good arm so what then is he doing, dropping everything? There should be a glove rating as well in that case, because otherwise it makes no sense to me how he can have so many errors.
Bob
November 02, 2002 at 12:49PM View BBCode
It is damn scary how much time we spend on fictional players -- but what the hell, it's a lot of fun.
I don't think errors are over-weighted, but I do think the defensive model needs to be revamped. To me this is the most important upgrade we could have.
The most glaring example of a problem is the one geoffrey mentions -- 1B. In SD, first basemen are always at the top of the list of errors committed because they routinely drop throws. Of course this is silly -- even a 1B with the range of a rain barrel (*cough* Mo Vaughn *cough*) can catch throws. Range should impact the number of plays a 1B can make, but that wouldn't show up in his errors -- it would show up in how many hits the pitchers give up and possibly how many throwing errors the other IFs make, since the low-range 1B wouldn't be able to dig out errant throws.
Another problem area is OF errors. In real life, most OF errors are due to errant throws. In SD, almost all OF errors are due to them dropping fly balls. While that certainly does occur, it happens too often in SD.
khakurat
November 02, 2002 at 03:38PM View BBCode
i just want to throw in my two cents on the allstar board.
while i think that geoffrey makes a valid point about guys who may not be getting to the ball, allowing a hit, and accumulating no errors, does a guy who has 33 errors really deserve to be an allstar? defense is half the game. i don't feel offensive abilities alone should determine whether or not a player is an allstar.
this might be tough to do, but i think a player should be listed on the allstar board by the position he's playing, not the position he's for which he's listed.
i really can't stand it when people don't actively manage their teams. to just administer to the basics doesn't really take that much time. that said, i feel that they should be penalized by not allowing their players to be eligable for the allstar team. college players get penalized for the infractions of their schools all the time, i just think it's fair to reward people who are spending the time, and penalizing those who are not. besides, it's not like the players are concerned about it.
finally, the cy young, fireman award, mvp, and rookie of the year should be there for both leagues, not just one each for the combined league. that's what they do in the bigs, so why not here?
there you have it, my 2 cents in 7000 words or less.
jer2911
November 02, 2002 at 04:34PM View BBCode
I think that back in the '50s, there was just one Cy Young and MVP for both leagues. (...But I could be wrong) I also don't think that they even had the Fireman award, and if it was there, the ROY award was still in it's infancy. Even so, you can still tell who the best pitcher was in each league by looking at the top of the All Star page and seeing who the starting pitcher is.
tysok
November 02, 2002 at 06:23PM View BBCode
First off, Jackie Foxx, even with errors out the wazzu, is still the highest ranked 1B in the game. On the same note, the point system is only subtracting the amount of errors, it's not weighted at all... I think that's more than fine.
Next Gilbert Robinson. He definately should be on the team... in fact a starter. I gave the run down in a post above somewhere... The top 4 2B in the majors look a lot alike on the base. Robinson scored the same amount of runs as the other 3, save Stennett who scored 20 more times than anyone. Even with 150 less ABs (around 600 ABs for the other 3) he still knocked 18 2B (32, 32, and 28 for the other 3), 33 HR (27, 15, 10) had 20 more RBIs than the 3, walked and struck out proportionately less, steal had the same SB as the other 3 (save Stennet with 53), but had about 4 times the errors (enter proper error problem argument here). But when you go down to the performance formulas, he has the same amount of Runs Created as the rest, his BPV is 30 pnts higher, and he had 16 Win shares which is about the same as the other 3. He did all this with 150 less ABs? If he were healthy all year he's an MVP candidate.
I've looked through some more and I would agree with Bob for the most part. The system in use now may be good enough. It's not as exact, it allows for a useless player that excels in one area (speed) to be taken above a great player that does everything equally well... but the arguments can be made either way. The biggest problem, and the one I've been looking atthe most, is the backups... if I had just realized what it was doing before I spent time looking into this, the whole post would be shorter. :)
Some info that changed my mind on that... I took the top 15 players in 4 different versions of formulas. The fantasy pnt that we use right now, the 4 pnt system I mentioned yesterday, also a 3pnt system (only fantasy, RC and BPV added), and a 2 pnt system (just fantasy and BPV added) and looked at who showed up.
While the top 15 did a little juggling around in each formula, they did all stay in the top 15... save 1 guy from each league.
First the AL, where it chose to put Al Zarilla above Danny Ferrer. Ferrer played less but did more in my opinion. Zarilla stole about 20 more bases, and the sstem was just a few pnts apart... lending to the idea that speed is weighted too much (which it really is but doesn't matter). However that's a matter of opinion, who was better? I think Ferrer was better, but the numbers can definately say to someone else that Zarilla was better.
In the NL it chose to put Steve Etchebarren above Bobby Thomson. There is no question that Thomson was better... he kicks Etchebarren's **** in everything but speedrelated categories. Etchebarren stole 41 bases while Thomson stole 0. Speed is definately what put Etchebarren up there. Take a look at the stats, Etchebarren plays for St. Louis, Thomson plays for Chicago NL.
Anyhow the conclusion is that speed needs to be toned down a bit... maybe 1.5X instead of 2X... but I don't think either of these guys would have made the team anyhow, so it could be a mute point.
We definately need to stop forcing a backup for each position, just get the best players that weren't starters. Other than that it may be fine, creating the controversial part of what is the All Star game.
tysok
November 02, 2002 at 06:27PM View BBCode
As for teams that don't get managed, the player is still good. It's the player we should care about whether or not he's an All Star, not the manager. Next season another manager takes over and the guy could get traded off...
I don't agree with forcing 5 SP and 5 RP, but that could be just the controversial part. :) I would like to see it force 4 RP... and the 10th spot is taken by the best SP or RP that wasn't chosen yet, and it does work out occasionally where a RP is in the top of the rankings... by pure fantasy number there are 8 RP in the top 25 pitchers.
Pages: 1