Lets look at the NPC survival rolls in Book of the Estate.

We already know from the Children Survival Thread that Greg was aiming for 50% survival rate to adulthood and the historical tends to be closer to 70%. And that the current system doesn't serve that (unless you do what Greg suggests and stop rolling at 7, +1 if you wish to get 70%).

Lets correct the BotE with the above errata, and simply look at the survival of the adults.
From 21 to 45: 1/80 chance of death (1.25%)
So +5 years = 94% survive (of adults)
+10 years = 88% survive
+15 years = 83% survive
+20 years = 78% survive
+25 years = 73% survive

About 3 in 4 young adults reach 45.

After 45, the odds for death double to 1/40 (2.5%).
So +5 years = 64% (of the starting 21 year olds) survive
+10 years = 57% survive
+15 years = 50% survive
+20 years = 44% survive

OK, that is pretty respectable. 44% of the 21 year olds reach a ripe old age of 65. That number seems quite high in my opinion for a pre-modern, though.

For 65+ very old people, the Grim Reaper brings his Special Scythe: 4/20 (20%) chance of death. Yikes.
+5 years = 14% survive (30% of the starting population of adults died in 5 years! Or 2/3rds of the 65 year old survivors!)
+10 years = 4.7% survive
+15 years = 1.5% survive
+20 years = 0.5% survive
And this keeps happening, about 1/3rd surviving after each 5 years.


Taking this back to the children's survival, if instead of rolling until they are 7 and then not rolling at all until they are 21, when would you have to stop rolling as children and have the adult 1.25% for the rest of youth to get the 50% / 70% end result?

50% end result:
10% mortality rate up to 4: 65.6% survive
Using 1.25% mortality rate from 5 to 21: 53% close enough.

70% end result:
5% mortality rate up to 3: 85.7% (this, by the way, is comparable to worst rates in modern Africa)
1.25% mortality rate from 4 to 21: end result 68.3% survival. Close enough.

Although actually, most of the deaths happen during the first year or so of the child's life. During infancy. So another way of doing this would be to have a big spike right in the beginning and then a more moderate decline.

Worst rates in modern world are in Africa, about 170 / 1000 children in places like Chad. The Sub-Saharan African average is around 100 / 1000.

We can see from previous numbers that we ought to hit something like 35%/15% mortality and then have the adult mortality take over. In the interests of keeping things simple, I'd front-load this, since the vast majority of deaths happen when the child is an infant. And it is easier to remember: oh yes, this baby was born last year, so special rules apply.

So:
1st year: 35% / 15% / 10%
2nd - 21st: 1.25%
End result: 50.5% / 66.1% (a bit too few...) / 70.0% (Spot on!)

So if you wish to get 50% survival, have infant mortality to be 35% (double that of Chad...), and if you wish to get 70% survival, have infant mortality to be 10% (average for Africa). Then just roll as for the adults. Easy.

Note: Since the BotE has done away with grade of maintenance effects unless it is a famine, pestilence or plague, it doesn't mess up these numbers under normal conditions.

You guys kill me! Keep up the good work!


T.

Since I seem to be suffering from a touch of insomnia tonight, I typed out a small program that calculates the odds for survival and life expectancy for each year of a character's life, using different assumptions:

'Tweaked Book of the Estates'
(70% survival to adulthood, high infant mortality as I posted in the first post. Since the 50% survival to adulthood simply makes the pool of adults smaller but doesn't influence the chance of death, the life expectancy is the same. This applies to 'Vanilla BotE' as well, since the probabilities are the same post-20 years.)
Infant mortality (0->1): 10%
Child & Adult mortality (1->45): 1.25% (thus 70% survival rate to 21)
Elder mortality (46->65): 2.5%
Very Old (66 ->): 20%

Results show:
Age: The age group, for instance 0 means everyone from birth to their 1st birthday.
Survivors: Assuming 100000 newborns, how many of them are left in the age group?
Die during: How many die during the year?
Life expectancy: If you just turned [Age], what is your life expectancy (mean of all the same or older ages that people died in)?
(Note, due to the numerical accuracy and rounding to the nearest integer, there are some problems both with big numbers (100000*0.1 = 10001?) and of course small ones. Don't worry about it. I am actually calculating the numbers in float, but when turning them to integers, little errors crop up.)

Tweaked Book of the Estate
Age Survivors Die during Life expectancy
0 100000 10001 40.38
1 89999 1125 44.87
2 88874 1110 45.42
3 87764 1097 45.97
4 86667 1084 46.51
5 85583 1070 47.05
6 84513 1056 47.59
7 83457 1043 48.11
8 82414 1030 48.63
9 81384 1018 49.15
10 80366 1004 49.66
11 79362 992 50.16
12 78370 980 50.65
13 77390 967 51.14
14 76423 956 51.62
15 75467 943 52.10
16 74524 932 52.57
17 73592 919 53.03
18 72673 909 53.49
19 71764 897 53.94
20 70867 886 54.38
21 69981 875 54.82
22 69106 863 55.25
23 68243 853 55.67
24 67390 843 56.08
25 66547 832 56.49
26 65715 821 56.88
27 64894 811 57.28
28 64083 801 57.66
29 63282 791 58.03
30 62491 781 58.40
31 61710 772 58.76
32 60938 762 59.11
33 60176 752 59.46
34 59424 743 59.79
35 58681 733 60.12
36 57948 724 60.44
37 57224 716 60.74
38 56508 706 61.05
39 55802 698 61.34
40 55104 688 61.62
41 54416 681 61.89
42 53735 671 62.16
43 53064 664 62.41
44 52400 655 62.66
45 51745 1293 62.90
46 50452 1262 63.35
47 49190 1229 63.80
48 47961 1199 64.23
49 46762 1170 64.65
50 45592 1139 65.05
51 44453 1112 65.43
52 43341 1083 65.80
53 42258 1057 66.16
54 41201 1030 66.50
55 40171 1004 66.82
56 39167 979 67.12
57 38188 955 67.41
58 37233 931 67.67
59 36302 907 67.92
60 35395 885 68.15
61 34510 863 68.36
62 33647 841 68.55
63 32806 820 68.71
64 31986 800 68.86
65 31186 6237 68.99
66 24949 4990 69.98
67 19959 3992 70.98
68 15967 3193 71.97
69 12774 2555 72.97
70 10219 2044 73.96
71 8175 1635 74.96
72 6540 1308 75.95
73 5232 1047 76.93
74 4185 837 77.92
75 3348 670 78.91
76 2678 535 79.89
77 2143 429 80.86
78 1714 343 81.84
79 1371 274 82.81
80 1097 220 83.77
81 877 175 84.73
82 702 141 85.67
83 561 112 86.61
84 449 90 87.54
85 359 72 88.46
86 287 57 89.36
87 230 46 90.24
88 184 37 91.11
89 147 30 91.96
90 117 23 92.81
91 94 19 93.60
92 75 15 94.38
93 60 12 95.12
94 48 10 95.83
95 38 8 96.54
96 30 6 97.22
97 24 5 97.83
98 19 4 98.43
99 15 3 99.00


Now, we can compare that to the nice link Thijs shared in another thread about life expectancy during Medieval times in landowning families: http://www.hyw.com/Books/History/Fertilit.htm

Life Expectancy:
Males Born 1276-1300
Age Prospects Total
0 31.3 31.3
10 32.2 42.2
20 25.2 45.2
30 21.8 51.8
40 16.6 56.6
60 8.3 68.3
80 3.8 83.8

We can see that on average, the Book of the Estate gives life expectancies that are about 10 years longer for 0-30 and then the difference diminishes to 5 years for 40, and is about spot on for 60 and 80. This tells us that we need to increase the mortality during the early part of the lifespan. Having gotten that darn 70% survival rate to adulthood working at last, I am not touching it (extending Child to 21 so that it stays the same). Instead, I double Adults and Elders mortality chances (to 2.5% and 5%) and run the calculations again:

Pretzel Book of the Estate
Age Survivors Die during Life expectancy
0 100000 10001 34.33
1 89999 1125 38.15
2 88874 1110 38.62
3 87764 1097 39.08
4 86667 1084 39.54
5 85583 1070 39.99
6 84513 1056 40.43
7 83457 1043 40.86
8 82414 1030 41.29
9 81384 1018 41.71
10 80366 1004 42.13
11 79362 992 42.54
12 78370 980 42.93
13 77390 967 43.33
14 76423 956 43.71
15 75467 943 44.09
16 74524 932 44.45
17 73592 919 44.82
18 72673 909 45.17
19 71764 897 45.51
20 70867 886 45.85
21 69981 1749 46.17
22 68232 1706 46.82
23 66526 1663 47.46
24 64863 1622 48.08
25 63241 1581 48.70
26 61660 1541 49.31
27 60119 1503 49.91
28 58616 1466 50.49
29 57150 1429 51.07
30 55721 1393 51.64
31 54328 1358 52.19
32 52970 1324 52.73
33 51646 1291 53.27
34 50355 1259 53.79
35 49096 1228 54.29
36 47868 1196 54.79
37 46672 1167 55.27
38 45505 1138 55.74
39 44367 1109 56.19
40 43258 1081 56.63
41 42177 1055 57.06
42 41122 1028 57.47
43 40094 1002 57.87
44 39092 978 58.25
45 38114 1905 58.62
46 36209 1811 59.33
47 34398 1720 60.03
48 32678 1634 60.72
49 31044 1552 61.39
50 29492 1475 62.04
51 28017 1400 62.68
52 26617 1331 63.29
53 25286 1265 63.89
54 24021 1201 64.46
55 22820 1141 65.01
56 21679 1084 65.54
57 20595 1029 66.04
58 19566 979 66.51
59 18587 929 66.96
60 17658 883 67.38
61 16775 839 67.77
62 15936 797 68.13
63 15139 757 68.45
64 14382 719 68.74
65 13663 2733 68.99
66 10930 2186 69.98
67 8744 1749 70.98
68 6995 1399 71.97
69 5596 1119 72.97
70 4477 896 73.96
71 3581 716 74.96
72 2865 573 75.95
73 2292 459 76.93
74 1833 366 77.92
75 1467 294 78.90
76 1173 235 79.89
77 938 187 80.86
78 751 151 81.83
79 600 120 82.81
80 480 96 83.77
81 384 77 84.72
82 307 61 85.67
83 246 50 86.60
84 196 39 87.54
85 157 31 88.45
86 126 26 89.34
87 100 20 90.25
88 80 16 91.12
89 64 13 91.97
90 51 10 92.80
91 41 8 93.58
92 33 7 94.32
93 26 5 95.10
94 21 5 95.75
95 16 3 96.55
96 13 3 97.12
97 10 2 97.80
98 8 2 98.33
99 6 1 99.00

This is almost spot on. In fact, it is so spot on that I am starting to suspect that the Raid modifier +1 was intended to get exactly this effect (it doubles the probability of going to the second roll and hence the chance of death). I.e. that warfare would be part and parcel of Medieval life, and rarely did a year pass without some kind of low-intensity conflict, especially for the males of the landowning class. And the peaceful years would be balanced out by natural calamities. So the tweaked BotE would be 'unrealistically peaceful & devoid of disease and famine'. However, the above link states that they avoided the plague years precisely since they didn't wish to be skewed by the high plague mortality.

Still, I think I would go with the Pretzel version and simply abstract away most minor skirmishing. After all, I have better things to do than track each NPC whether or not they happened to be in the exact area through which the raid went! :P And then add modifiers when they participate in big battles and such.

So these are the mortality chances I'd use
'Pretzel BotE' -version:
Infant mortality (0->1): 10%
Child mortality (1->21): 1.25% (thus 70% survival rate to 21; Note, I extend the Child to 21)
Adult mortality (21->45): 2.5%
Elder mortality (45->65): 5%
Very Old (65 ->): 20%

Or if you prefer the roll instructions (D20):
Infant mortality (0->1): 1st roll 1-2, no 2nd roll
Child mortality (1->21): 1st roll 1, 2nd roll 1-5
Adult mortality (21->45): 1st roll 1, 2nd roll 1-10
Elder mortality (45->65): 1st roll 1, no 2nd roll
Very Old (65 ->): 1st roll 1-4, no 2nd roll

This is so awesome. :D

I remember doing the exact same calculations...
Ok, now with your odds, how would you do all the modifiers from BoE?

fg,
Thijs

For women, see:
http://nocturnal-media.com/forum/index.php?action=post;quote=18553;topic=2416.45;la st_msg=18555
and the following posts.

As for your question, Thijs, I am not sure. I am not sure if we even have good numbers for how warfare in medieval times skews the survival odds. Certainly it should depend on how the war & battles go? Although disease ravaging the camp is a bigger killer, in particular amongst the commoners.

Ok, lets come up with some that 'feel' right..

fg
Thijs

Guys,

This is a really great display of passion for the game and milieu, however, I fell the need to remind everyone that the game is based on a mythology of sorts — a mythology where knights like Branor the Brown live to 120 years of age and are able be beat a lost of King Arthur's knights in the lists—including Arthur. Lanceleot, Gawaine and Palamedes! We — and I do not example myself in any way — tend to want to explain scientifically (or mathematically) all of these variables, but at the end of the day we're trying to impose real-world metrics on a fantasy. We need to let the stories be our guide perhaps, more than the scientific record — excepting those instances where the stories leave real gaps in our knowledge that must needs be filled in order for us to build a game mechanic.

System such as these, accurate as we can make them, will never yield a 120 yr-old knight like Branor the Brown.

My two cents, for what it's worth.


T.

Well, with N=1.6 million samples, you get:
120 1 1 120.00
:P
(That is just the expected result, by the way. Once the sample size drops small enough, the assumption that you'd lose perfectly 20% each year breaks down. This just means that I get:
N* cumulative survival probability < 0.5 survivors for 121, so the last guy is labelled as a casualty.)

I am not worried about named, legendary individuals, whom I am quite happy to have outliers and exceptions to the rules. I wouldn't be rolling NPC survival rolls for Arthur, Guinevere, Lancelot, Gawaine and the rest of the Round Table. But the rank and file NPCs, family knights and so forth. And especially, the wives and children of the PKs, whose chance of death has a significant impact on the PKs' lives and stories.

Ok, lets come up with some that 'feel' right..


I am tempted to divorce battle & plague survival from the generic roll; just assign general percentages and roll separately. Kinda like the Family History does. For instance:

Was the NPC family knight at the Battle of Badon Hill?
Lets say that 75% of the Salisbury Knights were, so on a d20 roll of 6+, he was.

Did he survive?
Badon Hill is a murderous slog. I think Greg suggests killing 50% of the PKs, so taking that as the guideline: 1d20, 11+ he survives! (you could even tweak this: 10-11= He survives, but is crippled.)

Generic Battle:
Participation 1d20: 11+ he is there.
Death 1d20 (-1 if lost (-3 if against Saxons), -1 if hard-fought (Autun?)): 1 or less, he died.
(You could even put in 'captured for ransom' probability or on a win, 'captured a knight for ransom'.)

Same for disease. Generally, a famine wouldn't touch the nobility all that significantly. They'd have more to worry about if an actual peasant revolt is triggered, and then it is very bad news for local knights without strongholds. But I don't think we see those in GPC, at least not until the end? I'd assume that the generic wear and tear and common diseases are included into the survival roll already, so all we need to look out for is the actual epidemics, like the Yellow Pestilence. Then I'd just assign a probability of death, likely something between 10-20%, depending on the severity of the plague, and then roll. The first wave of Yellow Pestilence seems especially bad, so 20% seems about right. For the 'generic' plagues, I'd likely use something like 10%.

EDIT:
Of course, if the PKs make sure that a certain NPC knight is participating in the battle, then he is. No participation roll necessary.

I agree with you, but that would require quite a few extra rolls...

Is there way to combine them?

fg,
Thijs

I agree with you, but that would require quite a few extra rolls...
Is there way to combine them?


Ah, you mean the participation and death? Sure. The downside is that the participation is something that the PKs can significantly affect in game, so I'd rather keep it a separate one... But sure, let me see what I can come up with for those two examples...

Lets make Badon Hill participation 80%, because then we get a nicer probability of death. The End result will be:

Badon Hill (80% attendance):
1-4 = Didn't participate, survived.
5-12 = Participated, survived.
13-19 = Participated, died.
20 = Participated, died with Great Glory.
(Or if PKs dragged him along: 11+ = died, 20 = with Great Glory.)

Generic battle (50% attendance):
1-10 = Didn't participate, survived.
11-19 = Participated, survived.
20+ = Participated, possibly died. 1d20 again: 1-10 = captured and needs to be ransomed (if fighting Saxons, died), 11+ = died, 20 with Great Glory.
Modifiers: +1 if on the losing side, +1 if a hard-fought battle
(If PKs dragged him along: 20+ = died. (same modifiers))

The problem I see here is that the battle result modifiers skew the odds of participation, but I think it is a minor consideration? It is close enough. Ah, I figured how to solve it...

Generic battle (50% attendance, Modifiers: +1 if on the losing side, +1 if a hard-fought battle):
1-9 = Participated, survived.
10-19 = Didn't participate, survived.
20+ = Participated, possibly died. 1d20 again: 1-10 = captured and needs to be ransomed (if fighting Saxons, died), 11+ = died, 20 with Great Glory.
(If PKs dragged him along: 1d20 (double the modifiers): 19+ = possibly died. Roll 1d20 as above.)

This way, the battle results increase the chance of death if one participated, but do not change the participation rate. :)

Moving these rolls all the way into the survival rolls would be possible, too, but honestly, I think you are starting to make each roll too complicated at that point. Keeping the battle rolls separate from 'accidental death or disease' makes it easy to see what the guy died off. Also, it is easy to determine right there on the battle session, what happened to Uncle Aneirin? How many times do people roll during the normal battle anyway?

At the end of the Battle session:
GM: "OK, everyone rolls 1d20 for each their family knights. Up to 9 means your knight participated in the fight and a roll of 20 means he participated and may have died. Let me know if you roll 20."
*dices roll, most of the time NOT coming up as 20*
Player 1: "I got a 20!"
Player 2: "A 20 here too."
GM: "OK, reroll 1d20 for those two guys and let me know what you roll."
Player 1: "6. Is that good?"
GM: "He got captured and now needs to be ransomed." (Or if the GM doesn't wish to bother with that, the family knight's lord / comrades pile up the ransom, and moving on... Or say that since they won, they rescued the captives, no ransom needs to be paid?)
Player 2: "20 again."
GM: "He died with Great Glory. Make a note of that, and we can figure out something suitably heroic for the family history."

That takes what, 5 min, max, as everyone rolls a fistful of d20s and checks for 20s?

Generic battle (50% attendance, Modifiers: +1 if on the losing side, +1 if a hard-fought battle):
1-9 = Participated, survived.
10-19 = Didn't participate, survived.
20+ = Participated, possibly died. 1d20 again: 1-10 = captured and needs to be ransomed (if fighting Saxons, died), 11+ = died, 20 with Great Glory.
(If PKs dragged him along: 1d20 (double the modifiers): 19+ = possibly died. Roll 1d20 as above.)

This way, the battle results increase the chance of death if one participated, but do not change the participation rate. :)


Very nice solution!
You can even use this table for Badon with a +5 or so...

fg,
Thijs

Ok, lets make the rolls for plague and summarize..

Then we'll publish the 'Optional Book of Survival'..

fg,
Thijs

Ok, lets make the rolls for plague and summarize..

Then we'll publish the 'Optional Book of Survival'..

:D
I haven't followed the intricacies here
With your kind permission, if I can understand it I'll post it on my KAP website

Ok, lets make the rolls for plague and summarize..

Then we'll publish the 'Optional Book of Survival'..

:D
I haven't followed the intricacies here
With your kind permission, if I can understand it I'll post it on my KAP website


Thank, Greg!
Me and Morien would be honoured to have the labour of our mathematical explorations become part of the official Pendragon rules.
Perhaps Taliesin can help turn into nice downloadable pdf (and html table) that people can print and use during their games?

Of course Morien and I (and whoever else wants to chime in) have to agree... which can take a long time.. [scientists....]

fg,
Thijs

You can even use this table for Badon with a +5 or so...


Not quite. Because the way I made it, the modifiers only affect mortality in battle. However, you can do this:
If it is the Battle of Badon Hill, all results of 12+ are Participated, Dead.
(This gives 45% participated, survived; 45% participated, died; 10% Didn't participate, survived. Maybe a bit high participation rate 90%, but it is right next door to Salisbury and it is a do or die situation...)

Do I miss something:

Generic battle (50% attendance):
1-10 = Didn't participate, survived.
11-19 = Participated, survived.
20+ = Participated, possibly died. 1d20 again: 1-10 = captured and needs to be ransomed (if fighting Saxons, died), 11+ = died, 20 with Great Glory.
Modifiers: +1 if on the losing side, +1 if a hard-fought battle

50% participation, 10% casualty (of participators)

With a +5 the table effectively becomes: (a +5 on roll equals a -5 on table)
1-5 = Didn't participate, survived.
6-13 = Participated, survived.
14+ = Participated, possibly died. 1d20 again: 1-10 = captured and needs to be ransomed (if fighting Saxons, died), 11+ = died, 20 with Great Glory.

So 75% participation, 50% casualties (of participators)

Did I miss something?

fg,
Thijs

Did I miss something?


Yeah, you missed the part where the table got changed so that the non-participation was in the middle, to prevent the odds of participating shifting with the battle results.

Generic battle (50% attendance, Modifiers: +1 if on the losing side, +1 if a hard-fought battle):
1-9 = Participated, survived.
10-19 = Didn't participate, survived.
20+ = Participated, possibly died. 1d20 again: 1-10 = captured and needs to be ransomed (if fighting Saxons, died), 11+ = died, 20 with Great Glory.

Minor idea: add modifiers to also to second D20 roll [increase chance of glory/death]

Ok, how about 2 tables?

Generic battle (50% attendance, Modifiers: +1 if on the losing side, +1 if a hard-fought battle):
1-9 = Participated, survived.
10-19 = Didn't participate, survived.
20+ = Participated, possibly died. 1d20 again: 1-10 = captured and needs to be ransomed (if fighting Saxons, died), 11+ = died, 20 with Great Glory.

Important Battle
Generic battle (75% attendance, Modifiers: +1 if on the losing side, +1 if a hard-fought battle, +1 if long battle):
1-15 = Participated, survived.
16-19 = Didn't participate, survived.
20+ = Participated, possibly died. 1d20 again: 1-10 = captured and needs to be ransomed (if fighting Saxons, died), 11+ = died, 20 with Great Glory.

Note the +1 if long battle

PS: Out of my head, Badon is 3 day battle, so for 50% survival you need 20% 'died' per day of the participants.
Now if you have 75% attendance and 50% of those in attendance 'die', then you have 37%.5 (50%*75%) of all knight 'died'
That in 3 rolls means 85% survival per roll..
Badon is hard-fought and long, hence a +2 modifier, hence 15% on 'death'.
If however you want to catch that all in 1 roll

Badon (50% casualties, 80% attendance)
1-8 = Participated, survived.
9-12 = Didn't participate, survived.
13+ = Participated, possibly died. 1d20 again: 1-10 = captured and needs to be ransomed (if fighting Saxons, died), 11+ = died, 20 with Great Glory.

fg,
Thijs



PS: You could even just use the first (generic battle) table 3x, and consider only those people that roll 3x not-participated as really not at Badon...
[that means 87% of all knights fought at Badon]

Now we need odds for Raids...

fg,
Thijs

Yes, I was thinking of using the 3 battles separate rolling for Badon Hill, but I think it is simpler to have it just as a one roll.

The Important Battle Table works (although it should be 15-19 Didn't participate, if you wish 75% Participation rate). Badon just needs a note that it is exceptionally long (multi-day battle) and brutal, so +7 to the roll. This will result in 35% survivors, 40% dead, 25% not present, or to put it in another way, over 50% casualties of those who did fight at Badon. Which is an insanely high casualty rate for any victorious historical army to take. Most losing armies would have broken and routed already. But Baden is legendary for a good reason.

I am not sure you need the long -modifier. Any long battle is pretty much 'hard-fought'. Adding both will give the victor already 30% casualties (of the participating knights), which is pretty atypical. Even 20% tends to be a bit high.

Now I could accept that losing side might be +2 instead. I opted for +1, because I figured that the knights would be likelier to retreat in an orderly fashion or outdistance the pursuit being on horseback.


Your last comment about Raids:
We don't need it, minor skirmishing and such is already assumed into the normal survival rolls. Do you wish to track EVERY raid on Salisbury whether or not some family knight was present? I honestly don't. I might give them a roll if the PK drags them 'on camera' for one of the raids he is participating in, and just eyeball the result based on how well the PKs did (like 10% chance of getting unlucky if the PKs did well, 20% if the fight was hard for the PKs, 30% if the PKs lost and retreated after a hard-fought skirmish and 50% if the PKs were totally thumped).

Ok, one roll for Badon

I agree on the hard-fought and long overlap...
Both would be only for those famously long battles... (in GPC but not Badon).
I would agree on +1 for losing side, +1 if rout...
[most loses where orderly retreat]

Raids: You're right, only if dragged into it...
Personally I would like to write some more appealing raid rules where you can call upon your family/friends for attack and/or defense. [Rolling loyalty] And then some testing if something bad happened. Future/different thread.

But during anarchy you have those plunderings of ones land by saxons, surely you're family has more chance of kicking the bucket..
If your family knights are elsewhere, ok, but what if they stay on your lands?
What about your household knights?

What about bad harvests?

What about harse periods?
During anarchy I imagine more people were 'randomly' killed than during Arthur... hence the survival odds should be modified?

fg,
Thijs

I would agree on +1 for losing side, +1 if rout...
[most loses where orderly retreat]


You mean most losses when the losing side routs? Yeah, I agree with that. In fact:
Indecisive: +0 to loser (there is no clear loser)
Decisive: +2 to loser

It seems that generally, if you get a decisive victory, the opposing army routs. So makes sense that it would cause casualties. With indecisive, it seems that the opportunity to cause much more damage to the enemy is not there.



Raids: You're right, only if dragged into it...
Personally I would like to write some more appealing raid rules where you can call upon your family/friends for attack and/or defense. [Rolling loyalty] And then some testing if something bad happened. Future/different thread.


Yep, a different and interesting thread. :)



But during anarchy you have those plunderings of ones land by saxons, surely you're family has more chance of kicking the bucket..
If your family knights are elsewhere, ok, but what if they stay on your lands?
What about your household knights?


Normally, your family knights would NOT stay on your lands. If they do, they count as household knights and I hope you can afford them.

Depends on the scope of the raids.
Small pinprick cattle-rustling raids: included in the survival roll.
Fire-and-steel, make-a-desert: yeah, definitely increases the chance to die.



What about bad harvests?


What about them? Squeeze the peasants, live normally, peasants suffer as usual. The pnly time the nobility does poorly during a famine is when there is a peasant rebellion. Otherwise, they can take what they need. So what if another peasant family starves to death? Or a dozen? They probably would have died anyway, you know, before the spring...



What about harse periods?
During anarchy I imagine more people were 'randomly' killed than during Arthur... hence the survival odds should be modified?


Or Pax Arthuriana odds should be lowered? I think it is probably better to increase the chance of death during Anarchy than the other way around, as you do have tournaments, questing, robber knights and such during Arthur's time as well. Of course, those early reign battles would be extra.

Hmm. How about +1 modifier for those Anarchy Years when Salisbury as the whole is the target of Saxon raids (when tribute is not paid)? Then on a result of 2 and if the subsequent roll is death, you can state that the knights died fighting against the Saxons, while the women and children were kidnapped by the raiding Saxons (adventure hook)?

You mean most losses when the losing side routs? Yeah, I agree with that. In fact:
Indecisive: +0 to loser (there is no clear loser)
Decisive: +2 to loser


ok

Ok, for household knights the survival odds should chance if the land is raided/plundered/razed. Suggestions?

Hmm. How about +1 modifier for those Anarchy Years when Salisbury as the whole is the target of Saxon raids (when tribute is not paid)? Then on a result of 2 and if the subsequent roll is death, you can state that the knights died fighting against the Saxons, while the women and children were kidnapped by the raiding Saxons (adventure hook)?

Ok, so just include it in general survival.. ok

fg,
Thijs

Ok, for household knights the survival odds should chance if the land is raided/plundered/razed. Suggestions?


This has to do with the PKs' household knights, while the PKs are not present but the HHK is?

I'd be tempted to just say:
Raided: 5% chance that the HHK died defending the estate.
Pillaged: 10% chance...
Plundered: 20% chance...
That seems about right, eyeballing it, wouldn't you say?

If the PKs are around, then I might even roll for a singular HHK joining them and let the dice decide. If the PKs have multiple HHKs with them and engage in a fight against the Saxons, etc, I'd let the way the skirmish goes to decide the fate of the HHKs, like I mentioned earlier. I probably would diminish the dead chance a lot as long as the PKs survive and can first aid their unconscious/dying HHKs. And if the PKs fight to the last man, then I'd likely let the HHKs go down swinging with them.

I think that should be part of the future raid-rules...

fg,
Thijs