Rest of year update weekly changes

[alleyoops]

I lost track of the thread where this was discussed, but I believe the last I heard, the plan was for Todd and Gary to discuss which players had significant changes (e.g. playing time/role changes, additions, etc.) to the ROY projections in weekly articles. Looked around for articles related to the 5/23 updates, and can't find any. Are these identified anywhere?

[Todd Zola]

It was a hope, not a promise. I am going to try to do an NL review on Monday.

[Guest]

Im going to log an article probably for tomorrow's posting, maybe Wed on the process I've been going through in addressing playing time/performance issues.

Going forward I'll probably spotlight the significant team issues, though I can't say I'll address every playing time change in detail.

[alleyoops]

I realize that we're getting plenty of value from the site without this, and that it was not a commitment - rather a "hope", as Todd said. Still, since you guys are going through some kind of manual/analytical process to make these adjustments, it would seem to me to be fairly easy to include a few words to identify these kinds of changes.

The way that it was done on the full-season projections was fine, and probably overkill for my purposes. I don't need the reason - just some kind of identifier who was changed up and down manually. Since 95% or more players each week have no substantive change, it's like looking for a few needles in a haystack to see the "real" changes.

Thanks for your responses. If you can't supply this, no problem - just a request, not a demand.

[Guest]

I was going to keep all of this in site content but as a bit of a teaser here - here's generally how the process goes:

Each week a player's projected skill set is updated based on their performance to date. This doesn't change much, but for some it can and will (you see someone like Wang change a good amount already)

We take projected (from prior week) playing full season playing time and subtract time accrued to date and that serves as our baseline for playing time. We update for injuries/changes in role, then evaluate whether team totals for PA and IP by position make sense. So long as they do we probably dont need to tweak much more, if there is an outlier then we'll address it.

So when I say I want to do it on a bit more micro level - what Im saying is that every week, projections will be affected somewhat, but what I want to do is outline the areas where I had to manually interfere and did so with judgment that I am unsure about - for example how to allocate middle infield PA to Chicago White Sox IF or the 5th starter spot in NYY or SS in Boston, for example.

Anyways, will go a little deeper on this in plat content but this is the idea.

[Todd Zola]

Okay, I think I have a solution to how we can show players that have been changed. I am 99% sure I can go back and generate the data for the missing weeks, but I am 100% sure this could be a nice way of providing this info.

What we will do is have a separate worksheet in the RoY hit and RoY pitch general sheet that will show each player's expected total PA or IP based on that week's projection. It will basically be a total of the updated RoY AB or IP plus the year-to-date number. This way everyone can see on a week to week basis whose pace has been adjusted, and if desired, can ask about it in the forums if Gary or I do not happen to explain it elsewhere.

Does that make sense and if yes, will it work?

[alleyoops]

Okay, I think I have a solution to how we can show players that have been changed. I am 99% sure I can go back and generate the data for the missing weeks, but I am 100% sure this could be a nice way of providing this info.

What we will do is have a separate worksheet in the RoY hit and RoY pitch general sheet that will show each player's expected total PA or IP based on that week's projection. It will basically be a total of the updated RoY AB or IP plus the year-to-date number. This way everyone can see on a week to week basis whose pace has been adjusted, and if desired, can ask about it in the forums if Gary or I do not happen to explain it elsewhere.

Does that make sense and if yes, will it work?

That will work, as long as you include the previous week's projected AB or IP. Then we (or you?) could put in a calculation that highlights any player whose totals (AB or IP) changed by more than 4% from the previous week. It would be normal to see a 4% decrease in these on a weekly basis.

It would not identify any player whose AB or IP stayed close to the same, but whose value was tweaked (i.e. a skills adjustment). It would also miss a new closer (or one who lost the job), potentially, since their IP may remain about the same.

Perhaps a better method would be to show projected value for rest of year for last week and this week, with a highlight of those which changed by more than a few percent.

[Todd Zola]

What you will see is our updated end of year at bats or innings. If we never change someone, there will be the same number every week, matching what we gave in March.

You can them eyeball the number in the right-most column (the most recent updated number) with the number immediately to the left. If they are different, then we made an alteration. Those with some Excel acumen can write a formula to do it.

Letting people in behind the 4th wall, but this is information we need to input into the sheets to automate the process, so it is something we are calculating anyway. I just have to have a template with a vlookup to pull the number in. Anything more will lengthen the process and impact its timeliness. As I was doing the calculations for this week's update it struck me that we could provide this to hopefully solve this little issue. I'll just do the calculation after completing each week's numbers and not previous to the following weeks.

[Todd Zola]

The PA and IP have now been added to each RoY spreadsheet

[Guest]

Just remember that sometimes these things get done a day or two before they're posted and sometimes we'll miss something and not go back until the next week.

Not a cop-out, just don't always read into things unless you've asked here. Sometimes its just human error or we'll mis-evaluate an injury.