|
|
Post by mrsolidarius on Jan 11, 2022 4:53:06 GMT
Hello all! I'm new to the forum. I was fortunate enough to find you all when looking for a online solution for DBA 3.0. When I was young, I purchased a variety of Zvesda's plastic miniatures and also enjoyed the Total War computer series. I didn't know about wargaming back then, but I experimented making up my own rules to mimic the battles I had read about. I'm happy to have found an official channel for those of us who enjoy tactics and strategy.  I have a background in applied mathematics, so it seemed right to use that training for a good purpose. The result is a qualitative analysis on DBA 3.0 troop matchups: docs.google.com/spreadsheets/d/1JJel5e8AvFPZW2prRHZDi3CJ9Vtw7cyhLsCIx1k4VcU/edit?usp=sharingThe intent is to judge the strengths and weakness of element matchups in even number, advantaged, and disadvantaged states. This is done by simulating close combat under several different tactical factor combinations and counting the various combat outcomes that result (e.g., in a fight with no overlaps... solid blades win 8 out of 36 dice rolls against knights, cause the knights to recoil 13 out of 36 times, and are destroyed 15 out of 36 times). I then report the most likely outcome for that element matchup (e.g., solid blades typically lose against knights, meaning a wise general would probably avoid that matchup). Any feedback is welcome. I'm very new to the game, so I have much to learn, but I hope to create something useful or at least entertaining for fellow commanders. P.S. - I should mention that right now, I have only run the analysis for a limited selection of troop types. The document will continue to grow as I complete analysis on the remaining troop types. |
|
|
|
Post by kaiphranos on Jan 11, 2022 13:52:42 GMT
Very interesting! I remember trying a similar approach at one point to determine who was favored in Hittite Imperial vs New Kingdom Egyptian - basically made a grid with all the unit types in each army and compared each matchup based on combat factors and outcomes. Then I assigned numerical "strength" values to each unit type based on its performance against likely opponents and tallied these up to determine which was the "stronger" army. (My analysis gave it to the NKE, but that hasn't really borne out in practice so far.)
Anyway, I expect one could do something similar using this as a more sophisticated method for comparing the unit matchups - I'd be interested to see if the results matched players' expectations and experience!
|
|
|
|
Post by mrsolidarius on Jan 11, 2022 18:46:48 GMT
Agreed. I am also very curious on how the "theory" matches the reality of what players actually experience. I know for a fact that the theoretical performance of 4Pk against 6Cv does not match Cromwell's recent experience in his Swiss v. German battle! Pikes should generally perform well against knights, but in practice they were reported to feel "brittle," so that's definitely an interesting case study.
|
|
|
|
Post by chaotic on Jan 11, 2022 19:47:52 GMT
Any feedback is welcome. I'm very new to the game, so I have much to learn, but I hope to create something useful or at least entertaining for fellow commanders. Macbeth undertook a similar exercise. It would be worth having a look at: fanaticus.boards.net/post/3521 which identifies the factors he considered and the point structure he developed for the Magister Militum prize. My view is that the outcome provides a reasonable ranking of element types, but undervalues the effect of artillery, especially when more than one is fielded, and does not give sufficient weight to fast troops. However this probably just reflects my personal preferences. |
|
|
|
Post by mrsolidarius on Jan 12, 2022 7:18:32 GMT
Thank you for the source, chaotic! That's very much along the lines of what I'm interested in.
|
|
|
|
Post by gerardus on Jan 12, 2022 10:04:42 GMT
Some rambling thoughts on building upon what you have already done using examples against Elephants. Your conclusion of psiloi being weak against elephants. Whilst a common tactic is that you send your psiloi into elephants preventing it being directed against your other units that it can destroy. Psiloi will only recoil v elephants, so with the low psiloi base factor and the fact elephants will not recoil against the psiloi means once contact is made it will stick and will draw in the elephant in till it has overrun its support at which point you can deal with it with support of other units or push your advantage elsewhere. While there is always a small chance the psiloi will destroy it even on it own 1v1. Nuances like these are the challenge to modelling games like these. it could be done, but even when you do the analysis there are so many factors its often beyond (my) brains capacity to remember beyond some broad tried and tested guidelines when playing. Which for a wargame is probably a good thing as an statistician/mathematician does not always make the best general! To more accurately model this elephant example you would also need the following parameters added to your model, at the risk of overcomplicating a simple model... - Proportion of extreme risk advantage - So for example in countering elephants for example "kill El:to be killed by El" ratios for 1v1 encounters would be 3:0 for Psiloi; 0:6 for Bl; 0:0 for Sp+; 0:0 for Pk+ (proportion/ratio may be the wrong term here as numbers are out of 36 in total of course).
- Support Factors: Side support which benefits Sp+ in this example can be lost more easily than rear support which Pk+ benefits from here.
- Factor in the assessed disadvantage of troops that pursue (this is only by certain troops against certain troops). - Are there situations that pursue can also be an advantage?
- Your four conclusions: Counters, Strong against, weak against, countered by. Not necessarily assessed by the same parameters. Counters/countered by could be assessed by weighting your conclusion more on a measure of perceived extreme risk ie. what can it severely threaten/ what can severely threaten it . Perhaps weighted based on the Proportion of extreme risk advantage ratio outlined above, rather than solely on your adjusted most likely result. While strong against/weak against could be weighted more to your adjusted most likely outcome.
|
|
|
|
Post by mrsolidarius on Jan 12, 2022 20:09:46 GMT
Excellent example, gerardus! The notion of "extreme risk advantage" is a good one. An inherent limitation of the analysis I put together is that it does not factor in tactics beyond the immediate encounter of the elements. As you point out, psiloi may almost always recoil against elephants, but this does not necessarily mean they are a poor choice against elephants since they could act as a "shock absorber" preventing the elephants from engaging more vulnerable elements. You've listed a good possible way of incorporating this idea of "extreme risk advantage"... I'll have to ponder that for a bit. I should add that, I agree, a statistician/mathematician does not usually make a good general! There's a certain level of strategy that can only be gained by experience. That said, playing with numbers and applying them to a game we all enjoy is also it's own version of "fun"! And, as I hoped, I'm learning a lot from everyone's comments.  |
|
|
|
Post by medievalthomas on Jan 12, 2022 21:55:46 GMT
Pikes are weak against Knights because you have to double rank them so Knights always get an Overlap (so its +3! v. 4).
I have always used a market based approach for analyzing play balance. So when I did it for HOTT I gathered data from multiple tournaments. I looked at which Element types were the most popular. Which appeared most in the top three finishing armies and which appeared most in the bottom three.
This told me what players thought was effective (based on most popular). Which were in fact effective based on most appearing in top three armies. And which were ineffective - appearing the bottom three armies.
From this it became clear that Blades, Shooters and Riders were much more effective. Amongst 4 point elements Heroes enjoyed a clear advantage.
A similar approach to DBA 3.0 would probably get better results as it accounts for all sorts of terrain and tactical nuances missed by head to head theoretical comparisons.
My son's senior science project was to design and implement a testing procedure to determine whether Blades were the most effective troop type. He did it by creating a set army of 6 Blades and then 2 of each type of three other elements. It was pitted against an army with the same 6 extras but a different main troop type. We then played about a 100 games. The Blade army prevailed by a substantial margin.
TomT
|
|
|
|
Post by mrsolidarius on Jan 12, 2022 23:19:06 GMT
That's a great direction, medievalthomas. I'll need to track down if there are any DBA 3.0 tournament results and armies published online. That's a cool science project! Out of curiosity, do you recall if the engagements were generally line formations, or was the deployment based on varying the main troop type? I imagine that an effective deployment strategy might vary based on the army composition. |
|
|
|
Post by medievalthomas on Jan 18, 2022 21:42:23 GMT
We just used the standard deployment rules. Each player deployed as they felt best based on terrain etc.
It was a real pain getting tournament results and army lists and took a couple years to get a large enough sample. But yes we probably should attempt to do this. One nuance in HOTT there are no army lists you players just choose the best types of elements and glue whatever figures they want on to a base. Army lists in DBA 3.0 would mute the results as you are forced to take a greater variety of troops.
TomT
|
|
