Raising vs limping (vs folding) small pairs in EP

Fooling with some filters and decided to compare the difference in value I get by raising and limping my small pairs in EP. Keep in mind I'm a full-ring player.

Filters were:

-hands 77-22
-pf action = raise
or
-pf action = call and BBs called pf is <2 (HEM considers the first three positions "EP", and therefore calling pf could also mean calling a raise from earlier position instead of open limping, so I had to filter that out)
-minimum of 7 players

Raising:

482 hands, 17.80 bb/100

Limping:

274 hands, 92.72 bb/100

Folding should be considered too, but we don't need to look at HEM to know that we neither make nor lose money by folding. 0bb/100 if you will. It's clearly the worst option since the other plays not only yield positive expectation, but very positive at that.

So obviously my winrate for limping is much higher than for raising, but there are some other problems that have me on the fence about this.

For one, even though this is from a 185k hand database, the sample size is still relatively small (one of the many reasons why sample size is everything). I don't know exactly how many hands we'd need to make my winrates converge enough to really start making assumptions, but there's definitely the possibility that my hand sample could mean nothing.

Even if the winrates are accurate, metagame issues could mean that raising is actually more profitable in the 'big picture' because it balances my range so well.

I mean if I were to accept limping as a better play and eliminate raising with small pairs in EP from my game, that basically forces me to change my entire game to balance that play. I would have to start limping suited connectors and a lot of strong hands in EP in order to balance that. Then my raises with big hands become much more transparent because I will have 'moved' small pairs and suited connectors from my raising (or folding with respect to suited connectors) range to my limping range.

bah!

Not a very well thought-out post, just kind of posting my thoughts as I go. Discuss? I'm curious what you guys think.

I think for smaller stakes, we can easily throw balance out the window.

I've often thought I was losing value just stealing the blinds with my small pairs in 6max. But its hard to say.

Players at 25$ NL often stack off very light. So letting them see a flop, make TPNK, and stacking them with a set shouldn't be that hard to do. So idk, I've recently been sorta comprimising and raising less with my small pairs. I make smaller pot-building type raises (2.5xBB's) and get more callers. Not sure if that's a great idea either.

Thanks for the post. Shows me how much I still haven't even started thinking about.

No more THX lemur?

What limits are these hands from? I know sample size would reduce but I'd like to see the breakdown by limit. I think as you get higher, raising becomes the better option, for 2 reasons:

1. less people, especially in an unraised pot, will stack off with TPTK or other hands where bad players will pay off your set incorrectly.

2. Smarter players, they will begin to pick up on the fact that an EP limp means low pp set mining.

You mentioned #2, but I think #1 is the biggest reason, just because you are sometimes no longer getting implied odds to set mine, because you just won't get paid off enough.