We must have two threads on this subject. We play-tested this Saturday nigth and I have a combat simulation program I wrote. This is not like a 6+ armour save that might nothappen or benefit sometimes, it is between a 6+ armour save and a 6+ parrry save or a 6+ ward (like from the Luminark or ironcurse icon when hit by war machines). It is not nearly as valuable as a re-roll to hit, but it is like a partial re-roll to hit.
The right way to look at it is at the margin. It just gives you a bit more ability to ultimately win in a grinding match involving saurus, TG, saurus characters, CoC, or Krox with other large stubborn/steadfast units. If you have six wide ranks of TG in a big unit with a Slann, it is only another two attacks on average fromt the fronk ranks, but with buffs and S5 that can average an extra wound per turn. With the oldblood and scar vet kitted to survive combat and take few wounds and dish out wounds to higher T and AS models, the 6+ roll to hit extra attack really starts to add up in terms of killing stuff just that little bit faster. It may not sound like much, but it ends up mattering a lot over two and three rounds of combat. and that often can be the difference between winning, drawing or losing combat. If the war banner is worth signifcant points in the BRB and we see some ranked tournament players take it occasionally (maybe worth a bit less), then the ability to average +1 to ACR and also kill 1 model on average or put a fraction of a wound on a character or monster on aveage, then it is a valuable abiltiy. I have a simulation program and it looks to be worth maybe .5 to 1.0 points per model with saurus and a bit more on TG and more than that on Krox (50% chance of rolling a 6 and getting an extra attack and 50% chance of hitting and hit probability of wounding times average point per wound value ranging from 5 to 50 with an average of 8 to 10, then the value is about 1 to 2 points per model per combat phase). It becomes really valuable for Krox when fighting something with a high cost per wound like ironguts or monsters or cav models with high armour saves but lower T; then it becomes a huge benefit in combat. The same kind of value is realized by the kitted oldblood that can hit really hard in challenges and in fights with monsters and chariots and similar high T or high AS models that have high cost per wound.
In order to illustrate the value of PF on an oldblood, I ran some simple math analyses of an oldblood fighting a unit of empire models. and targetting characters or cav models with a great weapon. He hits on 3+ and wounds on 2+ and the armour save is only 5+. (This does not fully onsider the possibility of getting mutiple hits on multile 6+ rolls to hit). Five attacks means 5/6 of an extra attack (it is a bit more complex due to combinations), hitting 66.67% of the time times wounding 83% of the time and it is not saved 66.7% of the time. That is a 30.05% probability of an extra unsaved wound. If it is the captain on a mount with a dawnstone or 5+ ward save, then that is worth on average 17.3 points per combat phase. If it iis a typical inner circle knight, it is worth 7.7 points per round of combat. If it is a fully kitted war priest with a 5+ ward and on foot, then it is 18.5 points per combat phase. Against a unit of Demigryph knights, the oldblood will realize on average .30 extra unsaved wounds worth an average of 6.0 points per combat phase.
Another example is a hydra charging a TG unit with an oldblood or scar vet with great weapon. A total of 1 character and 3 TG will be in base contact (ignore the potential for supporting attacks getting extra attacks). The oldblood gets an average of 5/6 extra attacks and the TG get an average of 1 extra attack. The oldblood will cause on average .23 extra unsaved wounds and the TG will realize on average .104 extra wounds. The total wounds will, therefore, be .335 extra worth (even with the hydra under costed) 11.75 points. For comparison, typically the oldblood and 3 TG will cause 2.01 wounds on the hyrda without the extra attack, worth a value of 70.5 points, so the extra attacks increase the value by, not suprisingly, an extra 16.67% increase in value and wounds. The difference is more striking, however, when one considers two rounds of combat. Assuming the hydra wins or does not break in both rounds of combat (a likely result in the first round and somewhat likely in the second round of combat given that the hydra has a breath weapon (S5 given I2 TG and GW on oldblood), 7 cc attacks at S5, and a thunderstomp at S5 and 6 S3 AP attacks all at WS4.), then without PF the hydra ends up on average with 4.0 wounds after two rounds of combat and gets a third round in with a potential to do a lot more damage (decent probability the hydra gets the thunderfstomp in the second combat phase because the oldblood has an ASL GW) and holds up the TG through a third combat phase. On the other hand, with PF, the hydra suffers on average 4.67 wounds by the end of the second combat phase and more than half of the time is dead a the end of the second combat phase. You have a much greater probability of finishing off the hydra before the damaging thunderstomp in the second combat phase and a far greater probability of avoiding a third combat phase and the resulting damage and, if you do face a third phase of combat, you have a much better chance of killing the hydra before the third thunderstomp is resolved. Just to illustrate the damage a hydra does, after the first round of combat (no breath weapon and no more re-roll to hit), in each successive combat phase the beastmasters cause on average .67 wounds worth 9.33 points of TG, the hydra's combat attacks cause on averge 1.94 wounds worth 27.2 points of TG, and the thunderstomp causes another 1.94 average wounds worth antoher 27.2 poiints of TG. Thus, a full extra round of combat with the hydra kills on average 4.55 TG models. worth 63.8 points. The effective point savings on average attributable to having PF as opposed to not having PF is approximately 60 points equivalent to approximately 4 TG models saved. Remember that in the first round of combat, a hydra will killl a lot more with hatred and a breath weapon at S5, a total of 9.75 TG models worth a total of 136.5 points (not counting the champ which dies more than half the time in the first round when targetted with one attack from the beastmasters and then, if not killed, one attack from the hydra).
The range of outcomes varies dramatically depending on the army faced and the deployment and matchups the oldblood ends up in combat but an oldblood should get into combat a lot and killing a lot and realize at least a few poitns extra per battle even against basic infantry. That value does not consider the benefit of potentially swinging combats, more likely winning combat and breaking the oppsing unit (ocassionally causing +1 or +2 wounds from the oldblood and another +1 or +2 wound by the saurus or TG with him) and also taking away steadfast. This is especially the case when fighting monsters and cav and chariots where their ACR can be more variable due to thunderstomp, breath weapons and impact hit rolls causing a wide range of potential outcomes.
My son (top ranked LM GT player in US) and I agree that the extra attack is more like a re-roll to hit and the supporting attack limits do not apply since the attack is really only one and this is a bonus, not truly an extra attack like frenzy or +1 to attack always, and certainly should not apply with Krox in a skrox unit. We would expect an FAQ on this, since it is clearly the intent of the book author for this to aplly to all models in the unit.
