Evaluating Talent

Matrix Games Forums

Forums Register Login Photo Gallery Member List Search Calendars FAQ

My Profile Inbox Address Book My Subscription My Forums Log Out

Evaluating Talent

View related threads: (in this forum | in all forums)

Logged in as: Guest

Users viewing this topic: none

Printable Version

All Forums >> [Current Games From Matrix.] >> [Sports] >> PureSim Baseball >> Evaluating Talent

Page: [1]

Message

<< Older Topic Newer Topic >>

Evaluating Talent - 5/20/2006 8:34:14 PM

henry296

Posts: 418
Joined: 5/12/2006
Status: offline

When simming for more than a day, I see a message that Team X is evaluating talent which I think is a neat message. However, what does it mean when my team is evaluating talent? I'm not doing anything.

Post #: 1

RE: Evaluating Talent - 5/20/2006 10:44:57 PM

puresimmer

Posts: 2299
Joined: 7/24/2005
Status: offline

Your "scouts" evaluate talent periodically so if you ever ask for a suggestion or click auto-assign they will have some knowlege of how players are performing.

_____________________________

Developer, PureSim Baseball

(in reply to henry296)

Post #: 2

RE: Evaluating Talent - 5/20/2006 11:39:29 PM

VanScoy

Posts: 87
Joined: 12/2/2005
Status: offline

I have been asking this question since 2005! Thanks for answering it

. BTw Shaun, you have pried my OOTP06 money from me with this one. Keep it up!

(in reply to puresimmer)

Post #: 3

RE: Evaluating Talent - 5/21/2006 6:43:01 PM

henry296

Posts: 418
Joined: 5/12/2006
Status: offline

Shaun,

That's really neat. Can you give any insight in terms of "scout error". From the results, I know the scouts aren't 100% accurate in their ratings. Does every team have the same scouts so they all have the same error rate? Roughly how much range of error is there, +/- 10%, 20%, 30%?

(in reply to VanScoy)

Post #: 4

RE: Evaluating Talent - 5/22/2006 5:12:18 AM

lynchjm24

Posts: 240
Joined: 8/21/2005
Status: offline

quote:

ORIGINAL: henry296

Shaun,

That's really neat. Can you give any insight in terms of "scout error". From the results, I know the scouts aren't 100% accurate in their ratings. Does every team have the same scouts so they all have the same error rate? Roughly how much range of error is there, +/- 10%, 20%, 30%?

There is no scouting error. The ratings are 'real'.

(in reply to henry296)

Post #: 5

RE: Evaluating Talent - 5/22/2006 5:18:48 AM

henry296

Posts: 418
Joined: 5/12/2006
Status: offline

Thanks. There is just tremendous variabillity in how two 90 CH hitters will fare.

(in reply to lynchjm24)

Post #: 6

RE: Evaluating Talent - 5/24/2006 4:54:15 AM

henry296

Posts: 418
Joined: 5/12/2006
Status: offline

quote:

ORIGINAL: lynchjm24

quote:

ORIGINAL: henry296

Shaun,

That's really neat. Can you give any insight in terms of "scout error". From the results, I know the scouts aren't 100% accurate in their ratings. Does every team have the same scouts so they all have the same error rate? Roughly how much range of error is there, +/- 10%, 20%, 30%?

There is no scouting error. The ratings are 'real'.

I'd like to try to further understand the impact of 'real' ratings and see if my interpretation is correct especially since I see a lot of people saying the rely on stats instead of ratings. In my opinion since the ratings are real then ratings should matter much more than stats in terms of predicting future performance. If not, what am I missing? Let me explain my logic below.

Since the ratings are real then a 90 CH hitter is better than an 80 CH all things being equal (other visible ratings and health). Therefore, for every plate appearance the 90 CH hitter has higher odds of getting a hit than an 80 CH hitter. Is that correct? If so, other than rest there really isn't a point to pick the 80 CH hitter over the 90 CH hitter.

Now that doesn't mean you will always win more games with a 90 CH hitter over an 80 CH hitter because of the distribution of results may cause for a season the 80 CH hitter to outperform the 90, but the odds favor the 90 CH hitter.

Are there any hidden ratings, that would change the above conclusion.

Now most players aren't equal on all other metrics, but you are really making a choice between more contact, more power or a better eye when choosing between 2 players.

(in reply to lynchjm24)

Post #: 7

RE: Evaluating Talent - 5/24/2006 5:55:41 AM

Johnny Slick

Posts: 31
Joined: 11/2/2005
Status: offline

Since this game plays a lot like FPS: Baseball Pro... yes, a CH 90 player will be a better hitter than a CH 80 guy, all things being equal, but all things are never equal. The scouts are designed, I think, to look at the interplay between CH and PH and eye and speed and figure out from that who the best hitter in the best situation is. You can, as always, tweak the way the computer grades these abilities if, for example, you think it's overestimating power. Computer-run teams will also use age as a factor if they realize they don't have a shot at the pennant that year.

(in reply to henry296)

Post #: 8

RE: Evaluating Talent - 5/24/2006 6:19:33 AM

henry296

Posts: 418
Joined: 5/12/2006
Status: offline

I'm playing a 2005 replay season and it is currently June 2006.

I've got a player with 89 CH, 74 Power, 70 EYE

Last season he hit .254 and this season he is hitting .224 through 64 games. His OBP is under 300.

Should I view him as unlucky and that he just happens to be "rolling" outs instead of hits. Or is their something more to the engine and I should realize he will never be good and dump him because of his stats even if I only get a hitter with lower stats in all 3 categories.

What would you do with this player who is my highest paid player.

Thanks.
Todd

(in reply to Johnny Slick)

Post #: 9

RE: Evaluating Talent - 5/24/2006 9:54:47 AM

Amaroq

Posts: 1100
Joined: 8/3/2005
From: San Diego, California
Status: offline

henry -

I suspect Shaun would say he doesn't want to give away too much of how it works 'under the hood', and that figuring that sort of thing out is 'part of the challenge'.

For a long time - PS2003 through 2005! - I held the view you espouse in your first post, that the CH 90 hitter would always be better than the CH 80 hitter. I thought that the AI overvalued Power and statistical performance, and undervalued Eye, and won plenty of games and titles with a 'ratings first' philosophy.

However, I've also noticed players who look like they should get better performance than they do, consistently. For example, my current squad has a player who is Contact 91, Power 67, Eye 77, and has been very close to those number for four seasons: .277 .276 .257 .245 .. by the three 'hitting' stats, he should be a .300 hitter: I certainly have had a bunch of hitters post .290 and better with numbers worse in *every* major category.

Now, he does have a Speed of 39 - and if a number of ground balls turn into contests like "Fielder's Arm vs hitter's Speed", maybe he loses enough of those to keep his average down.

... but, experiences like that have been convincing me that maybe there's more going on than a simple, straightforward application of the visible ratings.

Maybe 'Contact', 'Power', and 'Eye' are overlays on top of a different set of ratings - a composite 'blend' of a more complex set of ratings that are used to determine the play outcome.

Maybe there are 'hidden' ratings that influence the results.

We can prove that there are hidden ratings in the game, period - if not that they're involved in a single at-bat. Have you ever noticed a veteran player who never recovers to 'Feeling Great'? He has a 'cap' on his 'maximum condition'. That 'cap' is not visible. Likewise, there's a 'players internal durability rating' which is referenced in the .xml file - either that file is lying, or there are hidden ratings. (Or are those the same thing?)

...

So... feel free to invent your own superstition, or attack it mathematically to come to your own conclusions!

All I know is, in the 2002 season I traded away two left-handed starting pitchers whom I was convinced were utter junk: despite good ratings, both were performing poorly for me. One has been utter junk ever since, truly one of the worst pitchers in the association... and the other has won the Cy Young award.

(in reply to henry296)

Post #: 10

RE: Evaluating Talent - 5/24/2006 6:36:10 PM

Amaroq

Posts: 1100
Joined: 8/3/2005
From: San Diego, California
Status: offline

quote:

ORIGINAL: Amaroq
However, I've also noticed players who look like they should get better performance than they do, consistently. For example, my current squad has a player who is Contact 91, Power 67, Eye 77, and has been very close to those number for four seasons: .277 .276 .257 .245 .. by the three 'hitting' stats, he should be a .300 hitter: I certainly have had a bunch of hitters post .290 and better with numbers worse in *every* major category.

Apparently he reads the forum, for after getting called out like that, he's now batting .327..

quote:

Maybe 'Contact', 'Power', and 'Eye' are overlays on top of a different set of ratings - a composite 'blend' of a more complex set of ratings that are used to determine the play outcome.

I forgot to mention that I think this theory is *almost* proven as well, simply by the fact that ratings are a T-scale. In other words, the ratings that are shown to us are a weighted scale, not an absolute scale. So, a '50' in power one season might not be a '50' in power the next season, and a player whose attributes are remaining constant could, in fact, have been getting better or worse in a league for which the average player was improving (or regressing) by the same amount.

What isn't implied from that is whether or not 'Contact' aggregates more than one 'real' attribute, or has a one-to-one attribution with a 'real' attribute.

(in reply to Amaroq)

Post #: 11

RE: Evaluating Talent - 5/25/2006 1:28:56 AM

KG Erwin

Posts: 8981
Joined: 7/25/2000
From: Cross Lanes WV USA
Status: offline

Amaroq, thanks for that info. Last night I composed a message about player ratings and how they influence draft selections. This was BEFORE you noted the non-granularity of actual player performance vis a vis ratings.

Yeah, you depend on your scouts to tell you about hot prospects, but in the real world, it's still basically a crapshoot, which is as it should be.

Anyway, here's what I wrote:
"
Whether using fictional or real players, everyone has a notion on the type of team they want to build.

Getting a grasp on the ratings and what they mean is an essential. Also, see this guide to handling the minors in PS2007: http://www.matrixgames.com/forums/tm.asp?m=1134126

When I first started, I concentrated on speed, pitching and defense. After studying the ratings and how they can improve, I realized that my obsession on managing every game was making me neglect my dual role as GM, and I had to really start thinking about who I drafted.

So, I want to discuss drafting philosophies. For pitchers, the essential ratings are in STuff and CONtrol. I tend to accentuate youth, so if I can find a young guy with an upward potential trend and high ratings in these categories, grab him up. For my 1946 real player draft league, Warren Spahn is THE prime example.

Batters -- in you want a team that can slap singles and move runners, the most important ratings are CONtact and EYE. These are the guys who will be patient and wait for a good pitch, and are likely to get a base hit or draw a walk.

SPeed is a mixing blessing -- high speed ratings will mean more scores from a baserunner on second with a single. However, unless the guy's speed rating is 90 or above, steal bases in a conservative way. Check out the opposing catcher's ARM rating.
In thinking about base-stealing, a CS does two things --it creates an out AND prevents a potential run. A double whammy.

Defensive ratings are another factor -- outfielders with great arms are likely to throw out a guy at the plate, BUT, if his RAnge and HAnds are low-rated, he might bobble the ball and let a runner advance or score.

Because of this, I desire my infield to be solid defensively. If you're lucky enough to get a good combo at short and 2nd, that helps prevent some singles from getting through the gap.

Compounding all of the above is tailoring your team to the ballpark you play in."

So, you blew a hole in my drafting strategy, but I wanted to offer my thoughts just to encourage further discussion.

(in reply to Amaroq)

Post #: 12

RE: Evaluating Talent - 5/25/2006 3:47:33 AM

Amaroq

Posts: 1100
Joined: 8/3/2005
From: San Diego, California
Status: offline

Draft strategy is a whole 'nother question!

But, I don't think this blows your strategy out of the water at all. You don't have any other data to go on, so you have to rely on the ratings - in fact, I pretty much rely on the ratings all the way up to the major-league level, and 'adjust' only based on what a player can do in the bigs - e.g., my 91/67/77 guy is still a reasonable major-leaguer, and I used him in the #7 hole; I call for 'hit and run' more frequently with him thanks to the high Contact, and I get plenty of use out of him.. he just isn't the star of my team.

In draft strategy, I, too, build a club around speed, pitching, and defense - and in fact, the entirety of your strategy is very similar to what I've developed.

I definitely 'filter' my draft selections around 'speed', 'hands', and 'range', (plus 'arm' for C, 3B, and OF) because I know those don't improve. The guy can have the best potential in the world, but if he's a 30-30-30 fielder, he's always going to be a major liability defensively, turning outs into hits, and bobbling some of the balls he does get to. I'm also pretty confident that those have a one-to-one ratio between the visible ratings and the outcome - compare fielding percentage to 'hands' for all the teams' starting shortstops at the end of a season, and you can see that pretty clearly.

Like you, I'd concluded that Stuff and Control are more important than Velocity for a pitcher; it feels to me like Control helps keep pitch counts down, so I was rating it pretty highly when I was a GM-only. Now that I'm managing every game, I've noticed that lower Stuff pitchers yield more home runs - my top-Stuff pitchers only seem to yield the big fly in high wind conditions.

But, for draft strategy, Stuff and Control will improve noticeably over the life of a youngster's minor-league career - Velocity won't (or not as much, anyways). I don't have enough draft-history with '07 to have a feel for what's important / what works yet, but I'm wondering if 'high Velocity plus high Potential' will be good - the Stuff and Control catch up over the kids minor-league career?

(in reply to KG Erwin)

Post #: 13

RE: Evaluating Talent - 5/25/2006 4:47:38 AM

KG Erwin

Posts: 8981
Joined: 7/25/2000
From: Cross Lanes WV USA
Status: offline

Ok, thanks, Amaroq, but here's where I differ from my earlier draft strategy -- I mentioned above that I started filtering for contact and eye-- you gotta have guys who can advance those runners and create runs.

Apart from actually managing, there IS an art to constructing a long-term winner -- to make an analogy to the wargaming world, battles are won by a proper application of combined arms.

The principle applies to a winning team -- the trick is in finding the proper combination of ALL the necessary offensive and defensive attributes. Concentrating on speed, pitching and defense isn't enough UNLESS you have an extremely dominating combination of those traits in your lineup -- you still gotta score runs.

All I can do is offer general suggestions -- there is NO "magic formula".

I know we went off-track, so maybe it deserves a separate thread.

OK, with real players, post-WWII era, reserve clause on, my current generalized draft strategy is this -- get four strong starters, a closer, and establish your batting lineup first -- I want a competitive team from the beginning.

The major league team is first priority.

After establishing the starters, only then will I look for two good pinch hitters (one lefty, one righty).

Following that is fleshing out the bullpen -- if you have four good long-endurance starters, then all you need are guys to hold the line until you can deploy your closer.

The last step is filling out the minors -- if you wait long enough, the new players get generated, and you can go after those young scrubs, who most likely won't last long anyway.

This strategy only applies to the Golden Age, where contracts are for life, and you, as GM, can release and sign players at will.

If you want to manage every game, dumping financials saves a whole lot of time (and headaches on contract negotiations with primadonna free agents).

< Message edited by KG Erwin -- 5/25/2006 4:56:04 AM >

(in reply to Amaroq)

Post #: 14

RE: Evaluating Talent - 5/25/2006 9:25:11 AM

Amaroq

Posts: 1100
Joined: 8/3/2005
From: San Diego, California
Status: offline

True enough - I'm not starting guys who hit 40-40-20 but are speedy and great fielders!

What I'm doing is weighting speed and defense into my equation, or more accurately using a lack of *both* speed and defense to remove players from consideration. I've got some slow players - but they can still field their position. I've got some middling fielders - but wow are they fast on the basepaths.

All of them have to be able to hit competently - and for that, I do use contact and eye, definitely.

(Though I do usually leave a spot for one middle-infielder who is an awesome fielder, and/or a catcher with a gun, even if that person can't hit very much, as defensive substitutions late in a close game.)

I draft-to-start very similarly to you: my last starting draft was three good long-endurance starters, a strong R1, and a closer, followed by about six selections on hitters, the fourth starting pitcher, then two more hitters; after that, it was two or three pinch hitters plus three or so bullpen pitchers, and then youth + potential for thirty rounds!

Oddly, the player I selected 60th made the Opening Day roster as a bullpen pitcher! He only pitched five games, but he was on the roster to start with!

(I acutally used to take 5 starting pitchers to open, and decided that was close to 'cheating', as the AI wouldn't do that at all, and my #5 pitcher would have been a #2 or #3 on most other teams in the league.)

(in reply to KG Erwin)

Post #: 15

RE: Evaluaing Talent - 5/26/2006 8:26:20 PM

leegra

Posts: 44
Joined: 10/15/2005
Status: offline

Amaroq, KG Erwin, et al--I, for one, have found your strategy discussions and suggestions to be really thought-provoking; in this thread and in numerous other PS forums. I always enjoy your contributions and insights--they encouage me to re-evaluate my own approaches to handling associations and teams...

I also enjoy yours, and others, comments and observations on their favorite associations--Amaroq's starting around 1901 with authentic rosters, ball parks, and photos; and KG's Golden Age of Baseball associations, incorporating his particular approach to the game. I just hope all of you will allow the rest of us to share your enthusiasm and hard work by posting some of your templates, permitting us to experience what you have talked about.

PadresFan has created an excellent central posting site, and perhaps you all could allow him to present your templates, which would be particularly pleasing to those of us who have been reading with much interest your forum posts...

Anyway, please keep sharing your strategies and insights; they are appreciated...

Lee

(in reply to Amaroq)

Post #: 16