Help

Forum >> Help >> Ratings vs Actual Stats  

Post ID	Date & Time	Game Date	Function
#16980	04/01/2015 5:44:16 pm		Ignore
Ace Joined: 03/30/2015 Posts: 40 Inactive	I've read a lot of talk about players ratings on the boards. Specifically SI and POT and how they progress or get hindered. How about actual player stats. Is anyone taking them into account? Do some of you have bad stats and good ratings or good stats and bad ratings? Just wondering.
#16982	04/01/2015 7:19:32 pm		Ignore
Seca Joined: 05/05/2014 Posts: 5198 Waterloo Dinosaurs Legends	I've goofed around in Excel trying to find correlations between player skills or combinations of players and actual stats. Some correlations are better than others, but even the best are fairly weak. (I just look at my own players, so sample size isn't big). So yes, player skills give you an idea of performance, but players can certainly overshoot or undershoot expectations. Personally, skills and potential get you on the team, but actual performance gets your spot in the line-up.
#16987	04/01/2015 10:08:15 pm		Ignore
Rock777 Joined: 09/21/2014 Posts: 9595 Haverhill Halflings III.1	I definitely look at the stats. Some players consistently under perform. Others consistently over perform. There are hidden variable for the players that we can't see. From what I've seen, I'm guessing that some of these hidden variables include Trajectory, Clutch, Steal Aggression, and Durability. So if I have a power hitter who keeps hitting into double plays, maybe he has bad trajectory and bad clutch. Just like IRL, I think you really need to watch performance in this game to judge your players. Granted I was only at LL-5 last year, but here is an example of a POT 11 guy who has been doing really well for me. Conversely, a POT 14 who is close to losing his job. I look at past performance on waiver players too. If he did bad at LL-1 I might still give him a shot. But if that 118 SI guy has been doing poorly at LL-6 I'm not touching him. I recently dropped this fully developed 118 SI guy. He was making way to much money for middling performance (plus a weak arm for SS). Updated Wednesday, April 1 2015 @ 10:09:01 pm PDT
#16989	04/01/2015 10:12:36 pm		Ignore
newtman Joined: 11/02/2013 Posts: 3343 Inactive	Steve has said in the past that there are hidden values (steal ability is one that has been identified), due to these values we can't see you would need a fairly large sample to come up with the rough correlation of SI to actual stats on the average player as well as the variance. I just don't have the sort of ambition to undertake that project, though I do have access to SPSS to run it if I had the data to run.
#17000	04/02/2015 8:03:37 am		Ignore
PrivateSnowflake Joined: 01/06/2015 Posts: 1166 Bloomington Thunder Legends	Like Seca said, potential gets you on the team. I look for that more in younger players. In looking at 27+ year olds, past stats are the deciding factor. I have 12 pot guys who anchor the heart of my lineup. I'd much rather have that 11 pot guy with a .360 OBP than the 13 pot guy who just doesn't seem to get it. I don't and shouldn't need to be patient with those guys. I'm playing to win. It's like real life. Some guys have the tools to succeed but for some reason or another they don't pan out.
#17010	04/02/2015 1:38:02 pm		Ignore
Seca Joined: 05/05/2014 Posts: 5198 Waterloo Dinosaurs Legends	I feel it is worth adding that we've been told there isn't regression to mean. So a player's actual performance depends on: - his skills (both displayed and hidden) - opponent skills - environment (ballpark, weather, ...) - random Since there isn't regression to mean, random can cause a player to over or under-perform their skill set. I think PrivateSnowFlake's post is well said. Performance after a few seasons is a good metric. But (IMO) you want to be careful you have a big enough sample size to exclude a bad stretch of luck.
#17012	04/02/2015 3:52:29 pm		Ignore
Ace Joined: 03/30/2015 Posts: 40 Inactive	Good Stuff guys... Thanks for the input! I have seen guys with big SI and/or POT that have below what I thought they would have in stats. I have also seen players with lower SI and/or POT that have higher stats. Thanks again for your input! And anyone else that wants to respond!
#17013	04/02/2015 4:25:00 pm
admin Joined: 01/27/2010 Posts: 4985 Administrator	Yes, the goal is to have a certain amount of uncertainty that just looking at the player skills isn’t sufficient to evaluate a player. A lot of online games have a much greater level of certainty in player evaluation, but I think that’s unlike the real world. Recent statistics are probably the best indicator of future production – although they can also be misleading with rapidly improving youngsters. Steve
#17016	04/02/2015 7:21:23 pm		Ignore
Rock777 Joined: 09/21/2014 Posts: 9595 Haverhill Halflings III.1	The uncertainty is one of the things that makes this game really good.
#17111	04/06/2015 5:17:33 am		Ignore
Meccanodonte Joined: 04/21/2014 Posts: 370 Inactive	I second Seca's analysis. I'd like to add, also, that POT (or SI, it depends on what you're weighing) could express in some weird and dysfunctional combination. I mean, some pitchers with very high defensive values, some pitchers who lack control, some hitters with high speed and no inclination for stealing, some hitters with power/speed combo but not enough contact skills. On the other hand, might be succesful, even with lower POTs, pitchers without dominating stuff but pitchability or hitters completely without defensive ability.

| |

Reply