ericbabe
Posts: 11927
Joined: 3/23/2005 Status: offline
|
It was indeed toned done quite a bit after release. We released originally with historically high disease casualties, and players hated it, so we reduced it and allowed for more ways to mitigate disease loss. I checked many sources for disease losses and I don't think it would be right to have disease affect every unit a little bit every turn: from everything I found, disease hit particular areas over a short time, and it him some units very hard. I found regiments that took 60% losses in a single month, other regiments that would go years without reporting significant disease losses. Since I like to read myself write, I'll share my opinions on the whole game/simulation thing.... I say: "Game" and "simulation" are just terms that denote use. A game is something you use to have fun; a simulation is something you use to solve a mathematical problem. A simulation is successful if its solutions fall within the tolerance for error of the mathematical problem. Both are subjective terms: "games" are fun or not with respect to the people playing them, Tic-Tac-Toe is not fun for me, and so it's not really a game for me; "simulations" are only simulations relative to particular mathematical problems. I worked for the University of Michigan physics department and wrote simulations for years, so my perspective on this is colored by that: if all you care about is whether or not an electron makes it through the solenoid or not, then your simulation doesn't have to tell you its velocity when it exits, and it's still a "good simulation." If all you want to simulate about WWII is that economic power leads to strategic victory, you can set up a RISK board and it's a good simulation of WWII within the parameters specified. A standard mathematical technique I saw used over and over in physics is "assume human beings are 1 meter radius spheres of water" -- gives good results in many cases, fails spectacularly in many other cases. Talking about wanting "a simulation" without specifying the problem(s) one is trying to solve is kind of meaningless to me. We find that players want games to simulate (1) roughly historical casualty levels from battles, (2) roughly historical battle sizes, (3) roughly historical temporal paces to the wars, (4) roughly historical mobility restrictions, (5) historical levels of economic/troop development. These seem to be the largest areas of concern. My personal standard when designing these things are comparable table games, as from SPI, GMT, AH, etc. -- that's my general level of "simulation is good enough." We can't make a model that simulates everything, so we concentrate on the largest areas of concern that people have, the areas of concern of comparable paper games. These are the "parameters of our simulation." Similarly, I tend to judge simulations based on aggregate results, not on extreme cases. If 90% of the electrons predicted to get through the solenoid get through the solenoid, then the model is good. There will be extreme cases in any model; I just went to a very interesting talk by a military guy at Origins on mathematical models in military simulations, and as part of that talk he even mentioned that any model will have extreme cases, that military models are generally judged by the sorts of problems they solve well, not by the extreme cases that one can find to break the model. When players find things to be unhappy about with our models, they generally point to extreme cases rather than aggregate results. This is OK, and there are always steps we can do to reduce the number of extreme cases when people point them out, but people should realize that in a big system like this there will always be some extreme cases. If we were to try to make a mathematically rigorous simulation of, say, brigade-level supply in the Civil War -- rigorous enough for which I could make an argument in a peer-reviewed physics journal -- then it certainly wouldn't be based on a couple simple little game rules, it'd likely be a system of simultaneous differential equations. "+1 supply if adjacent to friendly territory; +2 if on a rail-line" just wouldn't cut it. We don't get paid enough to build these kinds of models , and I'm not sure that players would put up with a manual filled with differential equations even if we did (the "matrix" in Matrix Games does *not* refer to the Gauss-Jordan reduction method.) Many people find our supply rules intolerably complicated as it is now. When our disease results were closer to historical levels, people didn't like it. One guy was arguing that the game would have a more historical "feel" if the disease levels weren't at historical levels... in none of the accounts he'd read did military commanders ever announce they were taking disease casualties into account when they were making strategic decisions, and he was having to take disease into consideration when he was making his strategic decisions. Players who critique our model rarely make mathematical arguments ("Here's the distribution of casualty percentages for every major battle of the Civil War in comparison to FOF's distribution of casualties."), they're more aesthetic arguments, "Such-and-such doesn't feel historical." We take these concerns very seriously, but they are really more arguments about "fun" (historical feeling) than they are about simulations and solving mathematical problems. My conclusion will upset a lot of people, but it seems to me that when many people talk about wanting "simulations not games", what they're really talking about is having games with certain historical aesthetic qualities... just a different kind of fun really; they're not really critiquing a simulation in the way that a physicist or an economist would critique a simulation. I've been looking at a few boardgames lately that have gotten great acclaim from players for being "very historical" to try to figure out why people like them -- the systems they use are often very simple mathematically (draw your next reinforcement at random out of a cup), but they are designed to minimize the number of extreme cases, and they do this very well; these games are worse at solving mathematical problems, but they are better at preserving that aesthetic many people want: they are far less open-ended than other games, and the results they have are more controlled by the mechanisms of the game. I am considering using this as our criterion in future products since this technique seems to be well received when used for boardgames.
_____________________________
|