Originally Posted by
pbmax
Exponent, at this point getting very sloppy, used a master gauge adjustment to reconcile the two different readings the two gauges were giving, so that one set of numbers could be used to consider changes during the first half. Except that they used the adjustment for the halftime PSI readings and not for the pregame readings.
Update (9:45 p.m.): One of the issues in the halftime measurement of the footballs’s PSI is that two different pressure gauges were used, and one of them consistently came back with readings .3 to .4 PSI higher than the other. The Exponent report used a “master gauge adjustment” to be able to use readings from both gauges, and found that there was funny business going on with eight of the 11 balls. But Exponent made, according to Snyder, “a very basic mistake.” They used the master gauge adjustment for the halftime PSI readings, but not the pre-game PSI readings. If they had done so, they would’ve had very different findings:
Q. Let’s go to the next slide. And were you able to correct for that inconsistency that you described in Exponent’s master gauge conversion?
A. Yes. Now, the effective starting value is not 12.5, it’s 12.17.
Q. How do you get the 12.17?
A. You apply the master gauge conversion consistently to both halftime measurements, as well as the starting value.
Q. Okay. And let’s go to the next slide. And what is the impact of making that correction on the results?
A. Now eight of the Patriots’ balls are above the critical threshold predicted by Exponent, three are below.
So I have two studies that demonstrate that loss of pressure in this range is quite possible. In the study that purports to show that the loss cannot be due to natural processes, when errors are corrected, 8 of 11 balls meet the threshold.
And this does not account for the timing of measurement between the Patriots balls and the Colts, which might explain the other 3.
The evidence for deflation stinks and its not simply an opinion. If the NFL wants to rule on this matter, it needs better data and procedures.