Accuracy skewed by autohit?

[drdread]

I've been noticing that my controller has been missing more often. I started copying the "to hit" messages to an Excel file to examine my accuracy. In this case, I was facing enemies +0/+1 to me, but my accuracy was still stated as 95%. Now, my build has a lot of AoE and autohit effects, which I thought may skewer my accuracy. The enemies I faced were 5th Column, Council, Sky Raiders, Family, and Security Guards, so a mix of defenses, resistances and debuffs. Here is what I found from a sample size of 1635 hit rolls:

 

Result	Number	PRCNT
Hit*	546	67.6%
Miss	261	32.3%
Hit without autohit and streakbreaker	393	60.1%

  Overall, I had 774 autohits in 1635 hit rolls (47%). If I added all the autohits to my accuracy, I had an 81.7% accuracy. Closer but still significantly lower than the 95% the system reported I should have. Am I missing something here?

 

  The to-hit for the enemy was pretty accurate. The game reported between 50% to 66% to hit, and the finally average after 180 enemy to hit rolls was 57%.

 

I know the original randomizer that Cryptic made was not very good. They added the streakbreaker because the system would group-up hits and misses much too often.

 

PS. I can make the Excel file available, if that would be useful.

[Bopper]

A lot of powers, such as toggles, will not show a hit message nor a tohit roll. It may only show you a miss. When that is the case, the results will skew negatively. Now, the discrepancy you're seeing seems egregious still, so I won't say there is no problem. Just that it's worth keeping in mind that the combat log won't hold all the details you're looking for.