Trying to calculate pool tendencys

Hi guys,

I'm doing a database review over 200k hands on 25NL Rush&Cash 6max on GGpoker.
To practice math & get more skilled with using PT4, I've decided to not use the report tab, but calculate certain pool frequency's manually.
Starting off easy (I thought lol) I'm looking at RFI frequencies trough different positions.

I tried the following method (for looking at RFI % pool utg):
- total hands hero mp/cu/btn/sb/bb = 167,461
- total hands first raiser position UTG = 31,593

31,593 / 167,461 *100 = 18,8% RFI

I figured this was a legit frequency at 25NL zoom, only until I used the same method for figuring out RFI MP, which was to my surprise a lower percentage than UTG.
- total hands hero utg/cu/btn/sb/bb = 167,213
- total hands first raiser position MP = 29,007
29,007 / 167,213 *100 = 17,3% RFI

I can't figure out what I'm missing, can't imagine the RFI from UTG is actually higher than that from MP.
I tried to play with the no limping occurred filter, but that doesn't seem to make a lot of difference either.

For the 2nd example I used the following filters, simply put, I divided the value/quantity of the first filter by the quantity of the second filter:
TOT hands hero mp excluded:

TOT hands MP RFI