Now, let me pose a question to you: Assuming that Second Chance is reserved for critical hits until some time late in the encounter, how does this change how you evaluate it, mathematically?
Finally got back to this question.
As we discussed earlier, Second Chance should be saved until either the first critical, or until about the middle the encounter (rough rule of thumb is that 25% of non-minion foes are down, or 50% of non-minion foes are bloodied), or until the PC is seriously wounded and threatened (e.g. PC is 2/3rds damaged).
Assuming a critical does 150% and there ia a 50% chance of being hit, a re-rolled hit goes to no damage 50% of the time, normal 100% damage 45% of the time and 150% damage 5% of the time. Or, 52.5% of the damage instead of the 100% of a normal hit or 150% of a critical hit.
But, this is an "instead of" situation. For Second Chance, we have to calculate how much damage was saved, not how much damage was taken.
If he uses Second Chance, he will always average 52.5% damage taken. But, how much damage he saves is based on whether it is a critical or a normal hit.
On one hit, Second Chance saves 90% * (100% - 52.5%) + 10% * (150% - 52.5%) = 52.5%. Another way of looking at this is that an average hit does 90% * 100% + 10% * 150% = 105% damage and Second Chance does 52.5% damage, hence, the savings on one hit = 105% average - 52.5% taken = 52.5% saved.
On one hit, Shield saves 1-(1-0.3)^1 = 30%.
For two hits, there is a 10% chance that Second Chance stopped a critical on the first hit (97.5% saved) and a 90% chance that it stops 52.5% damage on the second hit. 10 * 97.5% + 90% * (90% * 47.5% + 10% * 97.5%) = 57% saved.
So, we solve for a range of N hits and how much damage is saved on average if the PC uses Second Chance or Shield on the nth hit (in the case of Shield, nth or earlier hit since Shield is used on the first hit it can be used on):
N 2ndCh Shield
1 52.5% 30%
2 57.0% 51%
3 61.1% 65.7%
4 64.7% 76%
5 68.0% 83.2%
10 * 97.5% + 90% * 47.5% = 52.5%
10 * 97.5% + 90% * 52.5% = 57%
10 * 97.5% + 90% * 57% = 61.1%
10 * 97.5% + 90% * 61.1% = 64.7%
10 * 97.5% + 90% * 64.7% = 68%
Shield saves 1-(1-0.3)^N.
We have to look at these numbers in the big picture:
1) This has slightly low numbers for Shield since Shield can stop multiple hits in the same round. Even if this only occurs one encounter in ten for a single extra foe, that increases the percentages for Shield by ~2%. Pro Shield.
2) Second Chance has a chance of not ever being used in an encounter if it is being saved for a later critical, especially in easier encounters. Combat is fluid and it's very possible for a player to think "Oh, it's stilll early, I should save it" and then combat is over in the next round or two due to good player dice rolls or area effects and the PC never gets hit again. Not being used drops the savings to 0%. Pro Shield.
3) A given player might use Second Chance on the third normal hit assuming that combat might be over soon and then get hit with a critical. One cannot just compare N = x between the two, one must realize that the Second Chance choice to use it has risks involved in using it too early. Using it too early on a normal hit drops the percentage down to 52.5% saved automatically, no matter how many hits were not used first. Shield should always be used the first time it can (unless the foe is known to be a wimp maybe). Pro Shield.
4) Second Chance has a better chance to save against specials such as daze because it can affect Fort or Will attacks. Pro Second Chance.
Taking all of this into account over the lifetime of a campaign, it really depends on how many times the PC gets hit per encounter.