Any opinions on this type of 80's player analysis?

[Sykes25]

Hitter and Pitcher files have been uploaded. Please see the other thread.

[bjs73]

[quote:e423ae88a1]For pitchers, I am done entering data. I add the ERA and WHIP together with the difference of Wins-Losses divided by 10 and Saves divided by 100. I then take that number and subtract it from their actual ERA and WHIP added. This discounts there actual win, loss and save total and adds some merit to how those pitchers performed on my team in terms of Wins. Losses and Saves.

For hitters, it is much easier. I do a straight up adding of the AVG, SLG and OBP and subtract those same numbers from their real numbers for that year. This help eliminate the AB issues even for subs or platoon players. [/quote:e423ae88a1]

Sykes, I love this kind of stuff. I spend all day at work trouble shooting quality systems, reviewing data, etc. This is right up my alley.

I'm going to take some time with what you've got so far (from the .pdf files) and see if the logic and formulas really give an unbiased picture of value. I've got some preliminary thoughts but I'd need some of my reference books to really try and dig into the meat of this project. And those are all at work.

My first thought in regards to pitchers is this:

If you truly wanted to add a useful "tilt value" for your pitchers (your wins/losses/saves formula), I'd dump the one you've got. But I do think you need to do something extra to validate pitchers though. This is what I suggest:

Keep the WHIP/ERA part of your formula and then bench mark their SOM performance against the league average. Booyah! That's the ticket right there. Let's look at Roger Clemens '88 Season (your first entry on your pitchers.pdf file) You've got him listed as -1.09 Value when you benchmark him against himself. Zoom out and look at the condition of your league he was in.

Go to the team stats page take the final league ERA (average of all teams) and the final league WHIP (average of all teams) and benchmark his performance against the league average.

And if you really want to get into whether a pitcher did well vs. the league average, calculate standard deviation of the pitching ERA + WHIP for each team. You'll get a bell curve like this one:

[url]http://www.clt.astate.edu/ub/Chemistry/Lab%201.htm[/url]

(This is a link to a Bell Curve that I like to use as a visual aid. The color one on this page is great.)

Figure out where Clemens falls on the bell curve based on your calculated standard deviation. If his performance fell into the green or blue area on the right side of the bell curve, he basically was a top performer in the league you were in, despite his individual stats.

Know what I mean?

Honestly, I think you could apply this concept in so many different ways to determine card value. What do you think?

[Sykes25]

Looking at league ERA is likely the easiest route to go.

When looking at value, I want to still take into account that although Clemens may have a negative value, he still is usefull. Just don't expect a 2.93 ERA every year.

All of this goes back to one of my original duscussions with Penn about looking at a player stat line to determine if he is a value for your park. My "value" now throws out the ballparks and takes into account what others do without having to directly correlate to their stats. Knowing that I have an ERA of 5.00 after game 42 or so on Langston means I have a likely 88 or 90 card, but if I am below that figure, I am likely to have 84 or 87.

I did not intend to develop an exact science to it. I needed to create a guide for myself to help show some results of all the darn teams I have played. :wink:

It's also a good tool for newer folks just to see raw stats of past performances since we do get new players to the 80's almost weekly nowdays. With spring training in a month or so, having this data ready to share is only a bonus to the 80's community.

[nycalderon]

I do agree that analyzing the data against the league average is helpful... but Sykes created a tool the he finds helpful to himself.... to understand how cards perform vs the actual data they are supposed to represent.

He included the won/loss tilt in order to compensate, I believe, for the fact that some good and effective pitchers get blasted occasionally in this game. A pitcher can have a slightly higher era or whip than you might expect... but the wins may indicate that he is in fact effective.

Certainly we could jumble the numbers around a million different ways but it's a great benefit to us that he wanted to share the data with us at all.

[BC Manager]

Sykes, I love the analysis, tools and the raw data you linked in the other thread. I think there's two flaws:

1 - There may be a skew based on the ballparks you used. As has been pointed out, I'm sure a disproportionate number of these seasons are from the Murph. So, the one's exceeding their real numbers are likely the ones who benefit from playing in that ballpark (ie., hitters who may have played in pitcher's park in real life have their numbers inflated at the Murph).

2 - The other issue I see with the data sample and consequently, with any analysis, is that you've only got data from [i:70e5b3e482]completed[/i:70e5b3e482] seasons. What about all the seasons for a player where they were crap for half a season and you dropped them. These stats aren't represented in this data.

Nonetheless I think it's all very interesting, and I already find myself looking up players I have on current teams are that I'm thinking of drafting, etc. Thanks for sharing. Awesome contribution!

[Sykes25]

You can't easily retrieve the data from dropped players during the year.

One important thing that I realized is that if a player managed to reveal 3 or more years to me at various times and yet I stuck with him until the end of the season, he is likely a player that I would not mind having regardless of what year may be hidden.

Kinda like that infomercial. "Set it and forget it!" :wink:

Tim wallach is a good example. Even in his worst years, he generally outhits his stat line. I can make a decision to stick with him knowing what I am getting for that price and move on to another position where I may not have that luxury yet. If money permits later on, I could revisit 3B if needed.

Yes, a lot of Murph teams contributed to those numbers, but a lot of Murph teams will be played again too. 8)