Jan 18 Projection BPIs

[Sean643]

I found some stark differences between some of the numbers in the HQ projections on the site and the ones in Rotolab.

Digging a little deeper I noticed that the stats seemed very similar but a lot of the BPIs were off (distinct differences in projected CT% PX SX etc).

I took a look at a simple example (Chip Ambres, 3BB, 5K).
Rotolab has a .6 Eye
HQ has a .73 Eye

A cursory look would lead me to believe that Rotolab is calculating the BPIs (correctly) rather than importing them and the site has some issues with some of the calculations.

I've tried to look to see if this has been mentioned anywhere and didn't see it (apologies if I missed it).

Is there something wrong with BPIs in the site projections? Surely 3 walks and 5 strikeouts is not a .73 batting eye (or have I been calculating it wrong?).

The differences in PX and SX mean that a Portfolio3 plan done with Jan 18 projections on Rotolab will look very different from one done using the CSV files

Am I missing something?

[RAY@HQ]

If you look at Ambres' Playerlink page, you see the issue: it's a discrepancy between the 'projected balance' and 'projected total' rows on that page.

I suspect the root cause is a rounding error, because it only seems to affect players who have very low BB/K totals. For instance, Erick Aybar also has his Eye skewed, but I just hand-calc'ed the Eyes for a bunch of other guys who are projected for more substantial playing time, and they're all either exactly right or within .01 of where they should be.

I should also mention that there's a major enhancement to Playerlink coming soon, which may also be a factor here.

So, I'll pass along to the techies, but in the meantime I'll suggest that the impact may not be as great as you think, because the effect is generally confined to guys with really low playing time projections.

[Sean643]

So, I'll pass along to the techies, but in the meantime I'll suggest that the impact may not be as great as you think, because the effect is generally confined to guys with really low playing time projections.

Ahh. Now that you mention it it does make sense as the largest discrepancies I noticed were with players very low ABs - I should have made the connection.

Thanks