O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

[huskerfan4life]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

@willard76 just shot ya a pm

[ShowTyme15]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

For those who may not know this is the best website for minor league transactions that happen weekly.

http://www.baseballamerica.com/t/transactions/

Also follow Matt Eddy on twitter.

https://twitter.com/matteddyba

[seanjeezy]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

I think I have a pretty strong formula for plate discipline, its something like 97% correlation: DISC = (668*BB%)+9

Good enough for what we are doing, especially after hearing the ratings aren't 100% objective in the last stream.

Here's one for vision, its not as strong (93%) but still good enough: VIS = (-269*K%)+115

I'm working on power atm, I think contact looks ok but I can check it out as well.

[CWSpalehoseCWS]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

I think I have a pretty strong formula for plate discipline, its something like 97% correlation: DISC = (668*BB%)+9

Good enough for what we are doing, especially after hearing the ratings aren't 100% objective in the last stream.

Here's one for vision, its not as strong (93%) but still good enough: VIS = (-269*K%)+15

I'm working on power atm, I think contact looks ok but I can check it out as well.

Interesting. How does this work along with minor leaguers at different levels?

[seanjeezy]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

Interesting. How does this work along with minor leaguers at different levels?

Using the projections and personal opinion, so for example DJ Peterson is projected for ~5.3% BB%, ~25% K% which would be ~44 and ~48. Personally I have him at ~52 and ~44.

If its a low level minor leaguer, it really depends lol. I have Alex Jackson at 30 discipline, 25 vision. That's something like ~3% BB% and ~33% K%

[Willard76]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

I think I have a pretty strong formula for plate discipline, its something like 97% correlation: DISC = (668*BB%)+9

Good enough for what we are doing, especially after hearing the ratings aren't 100% objective in the last stream.

Here's one for vision, its not as strong (93%) but still good enough: VIS = (-269*K%)+115

I'm working on power atm, I think contact looks ok but I can check it out as well.

We need to get yourself and Waittilnextyear working together on these formulas so we can get them zeroed in and applied to the spreadsheet

[CWSpalehoseCWS]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

Using the projections and personal opinion, so for example DJ Peterson is projected for ~5.3% BB%, ~25% K% which would be ~44 and ~48. Personally I have him at ~52 and ~44.

If its a low level minor leaguer, it really depends lol. I have Alex Jackson at 30 discipline, 25 vision. That's something like ~3% BB% and ~33% K%

Yea, that seems about right based on what I use as a general rule of thumb. This gives much more consistency though to get in the general ballpark, nice find.

[huskerfan4life]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

I was just wondering what are you guys going to do with the injured players like hunter pence out 6-8 with broken arm and yu darvish out indefinitely with a ligament tear in his elbow, will these players be on the DL or in A ball

[WaitTilNextYear]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

I think I have a pretty strong formula for plate discipline, its something like 97% correlation: DISC = (668*BB%)+9

Good enough for what we are doing, especially after hearing the ratings aren't 100% objective in the last stream.

Here's one for vision, its not as strong (93%) but still good enough: VIS = (-269*K%)+115

I'm working on power atm, I think contact looks ok but I can check it out as well.

You are dividing BB% and K% by 100 first, correct? If you keep it as a percent, the ratings are way out of bounds.

One question, have you done any sims? If your line shows a nice correlation, that's good, but it doesn't say anything about how accurate stats from these ratings would be in game. I would be interested in seeing what your league stats look like in some sims with these ratings.

I've already re-done knight's scales (so that they're based on rate stats rather than raw numbers--all normalized to 600 PA) based on some things we saw last year with making the Hybrid set (BBs originally being too low, SBs being too low etc).

Comparing your DISC ratings with mine, they're actually pretty close, but your top of scale guys and bottom of scale guys are a bit exaggerated from mine, which means your slope is steeper. For example, your Joey Votto DISC would be 133 using that formula and mine is 117. Your Trout would have 99 DISC and mine has 92 DISC. For the guys in the 50/60/70/80 region, our scales are quite similar. I am getting the inputs from ZiPS projections for this year.

As for your VIS scale, that formula produces ratings that are 5-10 points higher than what I have for guys over 50 VIS. For strikeout kings, our rating methods produce a similar outcome. I'd be reticent to raise VIS 5-10 points per player seeing what's going on in MLB as of late. Strikeouts seem to be only going up.

[RidinDwnKingsley]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

The one thing that I wanna stress is uniformity.... If we use the spreadsheet and ratings for 29 teams then we cant have one person rating another team completely different based on their own formulas. If you wanna collaborate and work on "formulas" together thats fine but in the end the ratings for all team will be uniform before released at least for version 1.


Sent from my iPhone using Tapatalk

[WaitTilNextYear]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

The one thing that I wanna stress is uniformity.... If we use the spreadsheet and ratings for 29 teams then we cant have one person rating another team completely different based on their own formulas. If you wanna collaborate and work on "formulas" together thats fine but in the end the ratings for all team will be uniform before released at least for version 1.


Sent from my iPhone using Tapatalk

Agreed.

My "formulas" are based on Knight's original spreadsheet but with slight tweaks for BBs and SBs. The formulas are just a graphical representation of what was already on Knight's chart but you can use Excel to make your ratings (according to the formulas) instead of wasting a bunch of time looking back and forth between stats and the chart in separate windows and eyeballing where in each bin that the guy's stats fit.

The tweaks came from what we saw from the sim engine last year while working on sim stats for the Hybrid roster. I prefer using Knight's stuff as a base to something brand new because it's been tried and true to work with the game's sim engine for several years granted the small tweak here or there.

[WaitTilNextYear]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

On another note....Contracts.

They are basically done and input into the spreadsheet. I am just going through 40-man rosters at this point and updating guys with no prior service time that are on the 40-man to "pre-arb 1" status. That s WIP, but will be done in a few days. All players above minimum wage should be in already. Feel free to double check your team! Please do! You have to scroll over to the right in the spreadsheet to see the contracts.

I wanted to get some feedback on a couple of things before moving on to something else.


1. If another team is paying a portion of a player's contract, do we want to reflect that? For example, reduce Matt Kemp or Prince Fielder's (there are a few others) contract to reflect the burden that only the current team is on the hook for? Do we not want to do that? We couldn't add money to the Dodgers or Tigers, but we could go half way here.


2. A lot of pre-arb players don't have salaries listed on Cot's Contracts yet, although that may yet happen before opening day. If not, I am thinking should we just use standard values for pre-arb 1, pre-arb 2, and pre-arb 3 players? Something like 510k for players with 0-1 years service time, 530k for players with 1-2 years, and 550k for players with 2-3 years. What do you guys think? Different teams have their own little accounting quirks for pre-arb players, but I'm not sure if deciding between 507k and 509k really warrants the level of research that would be needed to be accurate to the $1000 dollar for each guy.

[pimpinpidgeon8]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

I like the idea of giving players contracts they're team is actually paying them (Kemp, Fielder, so on). Now couldn't you add money to players' contracts on teams that are still paying money to players no longer on their squad? If that doesn't make sense, my apologies.

[WaitTilNextYear]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

I like the idea of giving players contracts they're team is actually paying them (Kemp, Fielder, so on). Now couldn't you add money to players' contracts on teams that are still paying money to players no longer on their squad? If that doesn't make sense, my apologies.

No, it makes sense, but would be creating a different problem. What if that player with the beefy contract gets traded? Then the acquiring team would have an unrealistic contract. For that reason, I'm not keen to do that. But, people could do that for their own personal roster if they wanted to.

[pimpinpidgeon8]

Re: O.S.F.M. 2015 Recruiting(joint ops with Hybrid Set)

True. Didn't think about that part. I agree with your take on the situation.