Blunder threshold in computer analysis

I actually prefer analyzing my games on my PC by some engine without even bothering using chess.com computer analysis feature. 

So, my question is about innacuracy and blunder thresholds used in computer analysis. Currently I use default values of 0.3 pawn for inacuracy and 0.6 pawn for blunder, but these values must depend on my rating (~1500 here at chess.com ). For example when I played at 900-1200 level I definetely would have used something like 1.5 for blunder and 1.0 for inacuracy.

Which innacuracy and blunder threshold values do you think are best suited in computer game analysis for differen rating groups (900-1200, 1200-1500, 1500-1800 ..) ?

I'm about the same rating. In an even game, I use about 0.3-0.6 for inaccuracy, 0.6-1.2 for mistakes and >1.2 for blunders. The more one-sided the game is the greater these numbers become. I also note blunders and mistakes that can't really be seen on my level separately

ryansth16, this makes sense, thanks for the reply. By the way which software you are using for automated annotation? I use Fritz Auto-Analysis feature, but always wondered maybe there is something better out there...

Sorry I didnt get back to you. I was using houdini but switched to Rybka since it apparently has a better positional algorythm. Is that a pay service? I dont know if I have an auto analysis feature so I also go largely based on principles and complexity as well. If I made a move that doesnt seem to make sense even if it only cost .1 pawn Ill make it an inaccuracy etc

I've started doing my own analysis of key games first where I think mistakes were made and then comparing that with engine output. I think I learn a lot better that way

I don't find the game analysis very helpful. Some moves are "Good Moves" like E2 to E4, who cares? everyone does that, thats not a "good move."  thats a standard generic move that everyone does.  When you call moves "good" that are just regular that doesn't help at all.  Then almost everyother move is an Inaccuracy,  of course it is, I'm a person not a calculator, everything is gonna be inaccurate always,  how is it helpful to point that out?  It should be you made a mistake or blunder or you didnt, no point calling inaccuracies.  If you do get the "best move possible" the analysis should let you know that.  Should be Blunder, Mistake, Good Move, Perfect Move, everything else should be neutral.  I mostly ignore the analysis and just read the final tally, and come to conclusions like "I won because the other player blundered 3 times and I only made 2 blunders,"  Also its just unclear what constitutes a Good Move.  Or what is an Inaccuracy as opposed to a neutral move.  How do you differentiate a blunder from a mistake? Just giving things names because you say so is not helpful.  Anyone who uses the Analysis should have a reference to these terms and how they are defined by the algorithm.  

                                                               ______________________________

It would be better if you stop almost all of your  BLUNDERS.  Or stop most of your BLUNDERS.

                          SIT ON YOUR HANDS

The extra second that it takes for you to remove your hand from underneath your leg in order to make a blunder move, will save you from many blunders.          

sitting on your hands is good advice but it's not the most necessary. i won an OTB 1 hour and 15 minute game with 10 second increment with my thinking time being 17 minutes (i've even factored in the increment.) while my opponents total thinking time was around 30-45 minutes. i wasn't sitting on my hands that game. 

Your spidy sense told you it was the right thing to do.

what do you mean by 'spidy sense' i'm not spider man. what is 'it' is it my 'spidy sense' telling me i should sit on my hands. or is it my sense telling not to sit on my hands. or was my sense telling me the moves i made to win the game.

spidy sense is from Spider Man.  Yes, definitely sitting on your hands is a very good idea.