Crits aren't that hard to take into account as long as you approach it with a particular range/multiplier in mind. Trying to account for every combo is where it'd get interesting, but the only difficult bit for a set crit rate is when you get to the point where you roll within your crit threat range but still miss.

Spoiler: Table Of Values, assuming 17-20/x2 crits
Show
x is your normal hit damage, assuming no bonus damage dice that wouldn't get multiplied on a crit.

Attack hits on... Chance You Miss (deals 0x) Chance You Hit (deals x) Chance You Crit (deals 2x) Avg Dmg
2 5% 76% 19% 1.14x
3 10% 72% 18% 1.08x
4 15% 68% 17% 1.02x
5 20% 64% 16% 0.96x
6 25% 60% 15% 0.90x
7 30% 56% 14% 0.84x
8 35% 52% 13% 0.78x
9 40% 48% 12% 0.72x
10 45% 44% 11% 0.66x
11 50% 40% 10% 0.60x
12 55% 36% 9% 0.54x
13 60% 32% 8% 0.48x
14 65% 28% 7% 0.42x
15 70% 24% 6% 0.36x
16 75% 20% 5% 0.3x
17 80% 16% 4% 0.24x
18 85% 12.75% 2.25% 0.1725x
19 90% 9% 1% 0.11x
20 95% 4.75% 0.25% 0.0525x


The problem with Weapon Focus - which is mentioned partially in the post that guy was responding to - is that, while it doesn't strictly make you worse at your job, it doesn't make you as good as other options do. But the problem is exacerbated by the existence of feat trees: Weapon Focus on its own is just kinda eh, and is vaguely comparable to some individual feats, but the Weapon Focus line is pretty awful and compares poorly with others. Since we're talking about a lvl 11 Fighter, lets use that as our baseline: this dude is getting 3 attacks with a weapon that deals x damage with a 17-20/x2 crit threat range (could any 19-20/x2 weapon with Improved Critical or Keen or whatever). Since we're being optimistic by expecting a competent Fighter with a nice attack bonus, let's assume that this guy's three attacks are hitting on a 4, 9, and 14 on the die (since that's in line with iterative penalties). That dude would have DPR of 2.16x. Throw on Haste (turning 4/9/14 into 3/3/8/13) and DPR is 3.42x.

Now, if he takes Weapon Focus, each of those hit numbers gets shifted up (to 3, 8, and 13), increasing DPR to 2.34x. Throw Haste on that (changing 3/8/13 to 2/2/7/12) and DPR is 3.66x.

If he instead took Rapid Shot, he's shifting all those numbers down 2 but is getting an additional attack at his highest bonus (so now instead of 4/9/14, it's 6/6/11/16), changing DPR to 2.7x. Throw Haste on that (for 5/5/5/10/15) and DPR is now 3.9x.

If he instead Power Attack, he could push his numbers down the track by n in exchange for changing {x} to {x+2n}. Let's crank the penalty to -4 (changing numbers to 8/13/18 and changing {x} to {x+8}. Now DPR is 1.4325x+11.46. Throw Haste on that (changing numbers to 7/7/12/17 and changing {x} to {x+8}) and DPR becomes 2.46x+19.68.

Here's a table comparing those options:

Feat Normal Is the best when x is... Haste Is the best when x is...
Weapon Focus 2.34x - 3.66x -
Rapid Shot 2.7x 9<x 3.9x 13<x
Power Attack (-4) 1.4325x+11.46 x<10 2.46x+19.68 x<14


And that's individual feats. Let's throw four feats each into the mix: Weapon Focus+Weapon Specialization+Greater Weapon Focus+Greater Weapon Specialization, Point Blank Shot+Rapid Shot+Weapon Focus+Dead Eye, and Power Attack+Improved Bull Rush+Shock Trooper+Leap Attack (ignoring charge bonus to attack). Let's also shift the basic assumption down to 5/10/15 to avoid cutting Slasher short.

Slasher turns 5/10/15 into 3/8/13 and turns {x} into {x+4}. DPR becomes 2.34x+9.36. Throw on Haste (for 2/2/7/12 and x+4) and DPR becomes 3.66x+14.64.

Shooter turns 5/10/15 into 6/6/11/16 and turns {x} into {x+7}. DPR becomes 2.7x+18.9. throw on Haste (for 5/5/5/10/15 and x+5) and DPR becomes 3.9x+27.3.

Pouncer takes a -4 to AC to turn {x} to {x+12} (not a small penalty, but not a large one either). Their DPR is 1.98x+23.76. Throw on Haste (for 4/4/9/14 and x+12) and DPR becomes 3.18x+38.16.

Also, for an additional category, let's take a person who's take an Int 16 Human Cloistered Cleric (with Divine Power for full BAB) and give them Knowledge Devotion+Item Familiar+Power Attack, full ranks in Concentration/the 6 monster knowledge skills/Spellcraft (with all skill points invested for +7 to each of the 6 knowledge skills), and the Collector Of Stories skill power. Now, for monster knowledge checks, you're effectively looking at +30, so you've got a great chance to always get that +5 Attack/Damage, giving you some wiggle room for PA tradeoffs. That build takes a -2 PA tradeoff to go from 5/10/15 to 2/7/12 and from {x} to {x+9}. Their DPR is 2.52x+22.68. Throw on Haste and increasing the PA penalty by 1 (changing it to 2/2/7/12 and to x+11) and DPR becomes 3.66x+40.26. Here's another chart:

Role Normal Is the best when x is... Haste Is the best when x is...
Slasher 2.34x+9.36 - 3.66x+14.64 -
Shooter 2.7x+18.9 20<x 3.9x+27.3 54<x
Pouncer (-4) 1.98x+23.76 3<x 3.18x+38.16 -
Knower 2.52x+22.68 1<x<22 3.66x+40.26 x<55

Shooter has Weapon Focus too, but only 'cause it's a pre-req; you could ignore the WF bonus it gets and it'd still be objectively better than Slasher. All four of these builds could do better by expanding their option even further, but it serves to illustrate the point of why people call Weapon Focus a terrible feat: it's not that it makes you bad, it's just that it makes you worse than you could be. When you take Weapon Focus instead of Power Attack or Rapid Shot or Knowledge Devotion, you're ever-so-slightly shooting yourself in the foot...but only a little bit. The problem is that the Weapon Focus feat tree is the obvious and intended bread-and-butter of every Fighter in existence, and the whole feat tree gives Attack +2/Damage +4. Shooter is better because it's getting Attack +2/Damage +7/+1 iterative, Knower is objectively better because it's getting Attack +5/Damage +5, and Pouncer is usually better because it's getting Attack +0/Damage +12.

When you take one feat that improves DPR by 10% instead of another feat that improves DPR by 15%...big deal, that's probably just a couple points difference in the end. When you take 4 feats that improve DPR by 10% instead of four feats that improve DPR by 15%, that's now a 20% difference, much more significant.