Cost efficient blood hunting

[chrispedersen]

Gut tells me its statistically significant. Haven't had to do such calculations in 20 years.. but it seems so.

[Tmoe]

Hello again! Unfortunately I was unable to edit my previous post so I have to write a new one. First off with the questions and notes.

Site frequency used was the Late Age standard of 35

As JimM stated earlier, B1+SDR or B2 will give you an average slave income per turn of 5 with roughly 10% of the time coming up short (rounding up). There after it's pretty much 1 slave per level of blood so the clear benefactor of the SDR is B1 hunter.

Indeed the stats show that using a SDR with 0-level blood hunters seems to backfire the results. As for the reason I have no clue.

As you all seem to love stas, here some more results

Next we have the blood slaves found/upkeep cost calculations. the first number after each entry refers to cost without the rod and the second with the rod. The entries listed here are more or less available when playing the game but I decided not to list subjects that have a 10% of getting the first or subsequent level of blood. Keeping in mind that the site frequency for testing was 35.

Early Age

Mitclan

Mitclan Priest 1,0 - 2,5
Priest King 0,9 - 1,1
Rain Priest 1,0 - 1,2
Moon Priest 1,0 - 1,2
High Priest of Sun 0,6 - 0,7

Sauromantia

Witch King (1lvl) 0,2 - 0,5
Witch King (2lvl) 0,5 - 0,6
Warrior Sorceress 0,4 - 0,9

Abysia

Warlock Apprentice 0,4 - 0,9
Warlock 0,5 - 0,7

Pangaea

Pan (1lvl) 0,1 - 0,2
Pan (2lvl) 0,2 - 0,3

Vanheim

Vanjarl 0,2 - 0,5
Vanadrott (1lvl) 0,2 - 0,4
Vanadrott (2lvl) 0,4 - 0,5

Helheim

Vanjarl 0,2 - 0,5
Hangadrott (1lvl) 0,2 - 0,4

Niefelheim

Jotun Skratti (2lvl) 0,3 - 0,4
Jotun Skratti (3lvl) 0,4 - 0,4
Gygja (1lvl) 0,1 - 0,3
Gygja (2lvl) 0,3 - 0,4
Gygja (3lvl) 0,4 - 0,4

Hinnom

Kohen 0,4 - 1,0
Melqart 0,4 - 0,5
Ba'al 0,3 - 0,4

Lanka

Raktapata 0,7 - 1,6
Yogini 0,3 - 0,7
Kala-Mukha (1lvl) 0,4 - 0,9
Kala-Mukha (2lvl) 1,0 - 1,2
Rakshasi (1lvl) 0,2 - 0,4
Rakshasi (2lvl) 0,4 - 0,5
Rakshasi (3lvl) 0,6 - 0,6
Raksharaja (2lvl) 0,6 - 0,7

Middle Age

Abysia

Warlock Apprentice 0,6 - 0,7
Warlock 0,4 - 0,4
Deamonbred 0,6 - 0,7

Pangaea

Pan 0,1 - 0,2
Pandemoniac 0,2 - 0,3

Vanheim

Vanjarl 0,2 - 0,5
Vanadrott (1lvl) 0,2 - 0,4
Vanadrott (2lvl) 0,4 - 0,5

Jotuinheim

Vaetti Hag 0,6 - 1,3
Jotun Skratti (2lvl) 0,3 - 0,4
Jotun Skratti (3lvl) 0,4 - 0,4
Gygja (1lvl) 0,1 - 0,3
Gygja (2lvl) 0,3 - 0,4
Gygja (3lvl) 0,4 - 0,4

Late Age

Ulm

Membr of the Second Tier 0,2 - 0,5
Fortune Teller 0,3 - 0,8

Marignon

Diabolist 0,4 - 0,9
Goetic Master (2lvl) 0,8 - 1,0
Goetic Master (3lvl) 1,0 - 1,2

Mitclan

Mitclan Priest 1,0 - 2,5
Priest King 0,9 - 1,1
Moon Priest 1,0 - 1,2
Sun Priest 1,0 - 1,2
Rain Priest 0,4 - 1,1
King of Rain 0,6 - 0,6

Abysia

Sanguine Acolyte 0,6 - 1,5
Sanguine Anathemant 0,8 - 1,1
Slayer Sanguine 0,8 - 0,9
Warlock Apprentice 0,6 - 0,7
Warlock 0,4 - 0,4

Midgård

Galderman (1lvl) 0,1 - 0,4
Galderman (2lvl) 0,4 - 0,5
Vanjarl 0,2 - 0,5

Utgård

Jotun Skratti (2lvl) 0,3 - 0,4
Jotun Skratti (3lvl) 0,4 - 0,4
Norna 0,1 - 0,3

Gath

Kohen 0,4 - 1,0
Kohen Gadol 0,4 - 0,5

Bogarus

Fivefold Angel 0,8 - 1,1
Occultist 0,3 - 0,7
Master of Names 0,2 - 0,5
Starets (1lvl) 0,1 - 0,3
Starets (2lvl) 0,3 - 0,4

----------------

This is just to give you an idea of how useful each entry is in blood hunting. However keep in mind that certain characters have issues that cant be taken into account. Good example would be Late Age Mitclan Moon or Sun Priests which suffer from old age.

Cheers

[Redeyes]

Once again, where is information that blood hunting depends on site frequency from? I also heard about it and probably it is so but is there anyone who can give a solid proof?

I would suggest that you enter two games, one with site frequency at 0 and one with site frequency 75. The difference in blood hunting results is easily evident that way.

Originally Posted by chrispedersen

Gut tells me its statistically significant. Haven't had to do such calculations in 20 years.. but it seems so.

From the series given, you can't meaningfully calculate standard deviation.

If you accept blood hunting at level 0 as occurring at a 10% succes rate, it isn't statistically significant (the results occur within two standard deviations).


And finally, to add to Tmoe's list:

Independents

Scout 0,3

[archaeolept]

a calculation as to your gut feeling? No.

it's pretty clear looking at the numbers for 1 lvl 0 hunter, there were 12 successes w/out the rod (out of 100) and ony 4 successes w/ the rod - when one would expect 10 hits, both of those are quite w/in normal expectations. The rod should not be affecting the chance of getting a result in the first place, and there has been no real reason to question this from these numbers.

the results w/ 2 sdr wielding lvl 0's were unlucky, and the results w/ 3 were about what one would expect.

but, if you wish to believe that giving level 0 blood hunters a useless item is harmful, feel free to not give them the useless item

[chrispedersen]

Originally Posted by archaeolept

it's pretty clear looking at the numbers for 1 lvl 0 hunter, there were 12 successes w/out the rod (out of 100) and ony 4 successes w/ the rod - when one would expect 10 hits, both of those are quite w/in normal expectations. The rod should not be affecting the chance of getting a result in the first place, and there has been no real reason to question this from these numbers.

Well, seeing as the rod *says* it requires skill in blood to use - AND the results to date support that conclusion...

I think its pretty clear that SDR inhibit your chances of success. But if you wish to continue giving these rods to people that can't use them - feel free.

Pending a larger sample size, of course.

[VedalkenBear]

If you want, a statistical analysis on if there is a significant difference in any of these numbers can be done.

[chrispedersen]

Originally Posted by VedalkenBear

If you want, a statistical analysis on if there is a significant difference in any of these numbers can be done.

Would you.. I would really appreciate it!

[archaeolept]

one point to make about efficiency, is that even unsuccessful lvl 0 bloodhunters cause unrest, so all in all you will be getting rather less blood per province set aside for hunting. There is a greater monetary opportunity cost per blood slave gathered w/ scouts than w/ actual blood mage blood hunters.

[JimMorrison]

Originally Posted by archaeolept

one point to make about efficiency, is that even unsuccessful lvl 0 bloodhunters cause unrest, so all in all you will be getting rather less blood per province set aside for hunting. There is a greater monetary opportunity cost per blood slave gathered w/ scouts than w/ actual blood mage blood hunters.

However, using Pangaea as an example, it's probably a worthy strat to start a couple typical hunting provinces with Pans, and roll all of those slaves directly into Empowering cheap indies to B1 and having them forge their own SDR. Done right you will need to be more aggressive about hunting to get a really solid income out of it (because of the cost of 53 slaves per hunter), but your gold/slave will be low, and once the machine is in place your blood income will be great, and all of your Pans will be free to do more exciting things.

I had never thought to treat a Blood nation as a non-Blood nation in such a way, I think it has the potential to reach a better middle ground than that of spending 70g/province in upkeep for the 3 Pans in it. Compare to the 2g/province upkeep for 3 Scouts with SDR.....

[Baalz]

And another angle to consider is the loss in income from the province, not just the upkeep of the mage. Particularly because the more blood hunting provinces you use, the less and less ideal they are - as you start wanting more bloodslaves you start bloodhunting in higher and higher population provinces. The more concentrated your bloodhunters are, the cheaper the bloodslaves (not even taking into account defensive considerations)

For instance, consider two different ways to go using LA Abyssia as an example. Lets say I'm bloodhunting with 2 B4 Abysian warlocks. Lets also say for the sake of argument, that not only do they have SDRs, but they each also have one blood booster which was forged for summonings/forgings and is not being used this turn, so they're effectively bloodhunting at B6. Extrapolating the trend shown in the graph (and that each blood level is supposed to gain you +1 slave) those two bloodhunters are going to pull in 14.6 + 6 = about 21 slaves per turn.

Alternatively you're using sanguine acolytes with SDRs because they seem on the surface to be more cost effective. Nice, cheap, holy blood hunters. Thing is, to get 22 slaves per turn you need 4 of them (plus the same boosters and an extra rod). Problem is, you can't put 4 B2 bloodhunters in one province or the unrest skyrockets. How's that upkeep difference looking as you switch extra provinces over to bloodhunting? This can be an enormous difference if you (for example) have a bunch of wastelands with a few high pop farmlands. You bloodhunting yet another 10k population province, or would you rather have your hunters spinning their wheels in a 3k mountain?

BTW, this is also the real cost in bloodhunting with scouts. Their upkeep is usually cheap compared to the lost income from the province they're bumbling around to eek out a handful of blood slaves.

Oh, also those two extra castle turns to crank out two extra bloodhunters (Warlocks vs acolytes) are not worthless resources.