Community Projections?

[ChiliIrishHammock24]

I'd like for us to copy what SSS does, and have a community projections thing for each player. Do a new player every other day, which means it would take 52 days in total (25 players x 2 days + division standings/awards x 2 days. Trust me, I got an A in Algebra).

 

So basically, we should start it like a week before pitchers and catchers report.

 

And it would look like this, for example....

 

http://spreadsheets.google.com/pub?key=t6V...amp;output=html

 

The way it works is you get a blank stat sheet to fill out, and then you submit your predictions, and it all gets added to a giant list like this. I don't think we need to delve into the standard deviations and all that, but I think having our predictions written down, and a community average would be cool.

[striker]

We should use my site www.fan-exchange.com. I store all of the info in a database and at the end of the year I use data from www.baseball-databank.org to score each individuals accuracy. I also upload zips, Bill James, cairo, chone, etc. so we can compare.

 

Here are 2010

http://www.fan-exchange.com/mlb/predictionsteams.asp

 

I have group competitions so it keeps soxtalk projections together.

 

All I have to do is update rosters and we are good to go.

Edited January 29, 2011 by striker

[ChiliIrishHammock24]

We should use my site www.fan-exchange.com. I store all of the info in a database and at the end of the year I use data from www.baseball-databank.org to score each individuals accuracy. I also upload zips, Bill James, cairo, chone, etc. so we can compare.

 

Here are 2010

http://www.fan-exchange.com/mlb/predictionsteams.asp

 

I have group competitions so it keeps soxtalk projections together.

 

All I have to do is update rosters and we are good to go.

 

Alright cool. But will it look like the SSS data page does, with everyone's predictions listed on one big page with a community average calculated out? That's what I most want to see.

[striker]

I can have it look however you want. Give me a week to update rosters and I'll get back to you.

[knightni]

I think that Joel McHale and Chevy Chase really carry the show.

[GoodAsGould]

I think that Joel McHale and Chevy Chase really carry the show.

 

Definitely agree with that

[ChiliIrishHammock24]

I can have it look however you want. Give me a week to update rosters and I'll get back to you.

 

Ok, that should be good timing anyway. I figured like around Feb-5th-7th is the time to start this.

[striker]

The community projection group has been setup.

 

1. http://www.fan-exchange.com

2. Register/Login

3. Click on Communities -> View

4. Click on Add Existing Community

5. ID = 15, Password = Peavy

6. Select White Sox as your team

7. Go add your predictions Predictions -> Stats (on left menu)

 

Let me know if you find any issues.

[Chisoxfn]

I'll pin this up for those that want to participate. JoeCool, you want to take the lead on this?

[ChiliIrishHammock24]

The community projection group has been setup.

 

1. http://www.fan-exchange.com

2. Register/Login

3. Click on Communities -> View

4. Click on Add Existing Community

5. ID = 15, Password = Peavy

6. Select White Sox as your team

7. Go add your predictions Predictions -> Stats (on left menu)

 

Let me know if you find any issues.

 

Does this require everyone to make an account to do the projections? Because I was hoping it could just be an easy thing like google docs where you just post a username and fill out the required stat fields.

[Chisoxfn]

I think the easiest would probably be to do it on here and somepone have a set template cause than no one else would need to re-register.

[ChiliIrishHammock24]

I think the easiest would probably be to do it on here and somepone have a set template cause than no one else would need to re-register.

 

I also want to do a less comprehensive list. Something like....

 

AVG, OBP, 2B, HR, RBI, SB, BB, K

 

W, L, ERA, WHIP, IP, K, SV

 

I think it would just be easier for people to fill out (and quicker).

[ChiliIrishHammock24]

Made this basic predictions chart, but can't figure out how to make the fields in a "response" form, and able for people to submit responses that are automatically compiled. I know it can be done, I'm just not sure how.

 

https://spreadsheets.google.com/ccc?key=0Ag...uthkey=COH36oEO

Edited February 8, 2011 by JoeCoolMan24

[chw42]

Why isn't there a category for SLG?

[ChiliIrishHammock24]

Why isn't there a category for SLG?

 

I figured it's too difficult of a category to guess, especially when you predict a 2B and HR, then your SLG could be way off what it should be based on that 2. Same reason I didn't include hits, because then there is math involved to get an accurate answer.

[witesoxfan]

OBP is going to be just as complicated. Preferably, what you would do is simply set up all of the counting stats and then enter in the formulas for OBP, SLG, and OPS. Actually seems pretty simple.

[chw42]

I figured it's too difficult of a category to guess, especially when you predict a 2B and HR, then your SLG could be way off what it should be based on that 2. Same reason I didn't include hits, because then there is math involved to get an accurate answer.

 

Or you could just include ask for the component numbers and get a script to calculate all the necessary numbers. That way, you can expand on those numbers into other more complicated stats if you wanted to.

[ChiliIrishHammock24]

OBP is going to be just as complicated. Preferably, what you would do is simply set up all of the counting stats and then enter in the formulas for OBP, SLG, and OPS. Actually seems pretty simple.

 

I will leave OBP, but eliminate walks. That solves that problem.

 

And the reason I didn't do what you suggest is because then for things like AVG, you have to put Hits and At-bats, and since people are more likely to have an AVG in their mind, they would just sit and tinker with it to get their AVG right. And then for SLG, I would have to put in 1B, 2B, 3B, HR, and AB. OPS would obviously be easy, but if I am not including SLG or OBP, then it's not applicable.

 

Basically, I just want to make this as simple for people, and as quick. I'm sure most people don't want to fill in 9-10 categories for 30 players on the list, so I just picked stats that don't have to make sense within other stats. I wanted independent stats (yes, I realize that RBI is somewhat independent of HR totals).

[witesoxfan]

If people are going to fill in 5 slots for 30 players, they are going to be willing to put in 10.

 

And if people have an average in mind, of course they are going to dick with AB and H to get it where they want to, that's kind of the point. It's pretty simple to do something like that. And for pitchers, I would actually reverse it and allow for IP and ERA and then figure out the ER from there.

 

For hitters - AB, 1B, 2B, 3B, HR, BB, K, R, RBI, HBP, SF

For pitchers - G, GS, IP, ERA, BB, H, K, HR, and SV too, though I think that will be pretty split between Sale and Thornton

 

For accuracy's sake, this seems to be the much better route.

Edited February 8, 2011 by witesoxfan

[chw42]

If people are going to fill in 5 slots for 30 players, they are going to be willing to put in 10.

 

And if people have an average in mind, of course they are going to dick with AB and H to get it where they want to, that's kind of the point. It's pretty simple to do something like that. And for pitchers, I would actually reverse it and allow for IP and ERA and then figure out the ER from there.

 

For hitters - AB, 1B, 2B, 3B, HR, BB, K, R, RBI, HBP, SF

For pitchers - G, GS, IP, ERA, BB, H, K, HR, and SV too, though I think that will be pretty split between Sale and Thornton

 

For accuracy's sake, this seems to be the much better route.

 

PAs would be better IMO. If I have enough relevant data, it's possible for me to convert these projections to WAR.

[ChiliIrishHammock24]

If people are going to fill in 5 slots for 30 players, they are going to be willing to put in 10.

 

And if people have an average in mind, of course they are going to dick with AB and H to get it where they want to, that's kind of the point. It's pretty simple to do something like that. And for pitchers, I would actually reverse it and allow for IP and ERA and then figure out the ER from there.

 

For hitters - AB, 1B, 2B, 3B, HR, BB, K, R, RBI, HBP, SF

For pitchers - G, GS, IP, ERA, BB, H, K, HR, and SV too, though I think that will be pretty split between Sale and Thornton

 

For accuracy's sake, this seems to be the much better route.

 

Why would them filling out 5 mean they are willing to fill out 10? That's twice as much, and even I don't want to fill out almost 300 spots. 300 and 150 are quite different.

 

"And if people have an average in mind, of course they are going to dick with AB and H to get it where they want to, that's kind of the point. "

 

I don't agree. I think a lot of people will say "I think Beckham will hit .290 next year", but then they would have to sit there and tink with hits and AB until they get a number that comes out to .290, and then then they do the 1B, 2B, 3B, and HR, they need to have it equal the hits they put in, ect.

 

Also, I don't see how HBP or SF are anything that anyone can guess with any sort of accuracy. Sure, you can assume Quentin will lead the team, and Pierre might be up there too, but those stats are too much dependent on the pitcher's inaccuracy, and aren't really numbers anyone is familiar with. Could you tell me who led the league in HBP last year or SF and with how many? Probably not. But you are much more likely to know who led the league in AVG, HR, RBI, SB, and maybe OBP.

[chw42]

Why would them filling out 5 mean they are willing to fill out 10? That's twice as much, and even I don't want to fill out almost 300 spots. 300 and 150 are quite different.

 

"And if people have an average in mind, of course they are going to dick with AB and H to get it where they want to, that's kind of the point. "

 

I don't agree. I think a lot of people will say "I think Beckham will hit .290 next year", but then they would have to sit there and tink with hits and AB until they get a number that comes out to .290, and then then they do the 1B, 2B, 3B, and HR, they need to have it equal the hits they put in, ect.

 

Also, I don't see how HBP or SF are anything that anyone can guess with any sort of accuracy. Sure, you can assume Quentin will lead the team, and Pierre might be up there too, but those stats are too much dependent on the pitcher's inaccuracy, and aren't really numbers anyone is familiar with. Could you tell me who led the league in HBP last year or SF and with how many? Probably not. But you are much more likely to know who led the league in AVG, HR, RBI, SB, and maybe OBP.

 

You can have people guess in ranges. That's what most projection polls do. They don't input exact numbers. Go look at how FanGraphs does it.

[ChiliIrishHammock24]

You can have people guess in ranges. That's what most projection polls do. They don't input exact numbers. Go look at how FanGraphs does it.

 

Well my goal is to find a community average, with a +5% above and below deviation to find the range.

[witesoxfan]

Why would them filling out 5 mean they are willing to fill out 10? That's twice as much, and even I don't want to fill out almost 300 spots. 300 and 150 are quite different.

 

"And if people have an average in mind, of course they are going to dick with AB and H to get it where they want to, that's kind of the point. "

 

I don't agree. I think a lot of people will say "I think Beckham will hit .290 next year", but then they would have to sit there and tink with hits and AB until they get a number that comes out to .290, and then then they do the 1B, 2B, 3B, and HR, they need to have it equal the hits they put in, ect.

 

Also, I don't see how HBP or SF are anything that anyone can guess with any sort of accuracy. Sure, you can assume Quentin will lead the team, and Pierre might be up there too, but those stats are too much dependent on the pitcher's inaccuracy, and aren't really numbers anyone is familiar with. Could you tell me who led the league in HBP last year or SF and with how many? Probably not. But you are much more likely to know who led the league in AVG, HR, RBI, SB, and maybe OBP.

 

Because 150 is an infinite amount more than 0. 300 is only twice that of 150.

 

Perhaps I just come from the school that if you are going to do something like this, you should try and find something of substance out of it. As it's currently constructed, you aren't going to get a great deal out of it.

 

If people believe Beckham will hit .290, they can also figure out about how many at bats they think he will get (as well as BB and HBP while making an educated guess towards SH and SF), and from there they can figure out how many hits he is going to get. They are already going to do doubles and homers, so you make an educated guess as to how many triples he ends up with, and from there you simply subtract.

 

The only reason you do HBP, SF, and SH is to help determine plate appearances, which will allow you to be able to calculate OBP, SLG, and then OPS.

 

If you are going to do OBP, you have to do SLG. Even if you don't want to make actual calculations based on doubles and triples and homers and whatever else, you freaking HAVE TO have SLG in there as well.

[chw42]

Well my goal is to find a community average, with a +5% above and below deviation to find the range.

 

You can do averages with the range method. You just need enough data points.

Edited February 8, 2011 by chw42