ATG Tank Models (v2.20)

[Tac2i]

Here is my first draft of trying to visualize this new feature:

1) Tank Model Diagram (same as below)

2) Tank Modeling Mind Map (updated 6/17/14)

Related Material:

1) Subformationtype Models (Vic's description)

2) ATG Mind Map

3) More mind maps

[Tac2i]

In a further attempt to make this new feature more clear I've created the below chart.



Update: per Vic there is a randomness built into this feature that means sometimes your designs turn out no better and sometimes even worse in a certain category. He also identifies a way to 'fix' this: http://www.matrixgames.com/forums/fb.asp?m=3712779.

[Josh]

Many thanks Roy. I agree with you that this new feature, although it is a great addition to this game, may need some ...uh let's say clearing up. Maybe the whole thing could be displayed in a more graphical style such as Hearts of Iron has. Sometimes when I need to produce some more tanks I loose the oversight what is exactly what. For instance I've created a new tank and it's called "Zebra" (as an example) when I go back to the production screen of the tankfactories I can't remember anymore what kind of tank Zebra really is, light, medium or heavy and with what modifications . Yes I know it is in the *description file* but still that should be somewhat easier to find out.

[Tac2i]

Probably the best thing to do at present is use some kind of naming convention that gives you a clue to what type of tank it is. That said,that may also give your opponent a better clue as well.

[baloo7777]

Thank you Webizen! Great work on the mind maps!

[Calliope]

Good work Webizen![:D] That makes it easy to understand.

But I think this new mechanizm still needs some improvements to fit the game better. It could be a good idea to improve your tanks but in most cases, it is not such a good choice to use PP in this area. The costs doesn't have such good returns like making factories, getting more troops etc. Both research and design are too expensive, which makes your less troops unable to hold your line. Also the upgrades are not that efficient and when the war starts it is difficult for everyone to put their production on PP for this TINY upgrade, and have to pay the high redesign fees. Furthermore, I have to say sometimes a bad design will ruin a game, especially in small map games the production is valuable and a little more troops will determine the result. What do you think after you put so much PP on it and then find that all your researches are just to make up the stats of a bad design? The risks should be satisfied with the returns, and in my opinion this mechanizm still needs more tests and improvements to make it work.[:)]

[Twotribes]

I don't use this feature as almost everytime I make a new tank it is worse then the stock tank.

[all5n]

I see this behavior also. It seems like the models are only based on the base class (i.e. Std. Light Tank), and not the currently researched level (i.e. Light Tank II).

Here is an example. I have Light Tank II researched, however when i create a new model (with Level II of all the upgrade tech), here is what it looks like:

Note that the model is inferior to the standard light tank II in many statistics. Is this working as designed?

[Vic]

Going to look into this. Will be back in a few days at most.

best wishes,
Vic

[all5n]

Some more info:

So after playing around with it a bit more, there seems to be a 50/50% for each tank upgrade tech to impart either "offensive" or "defensive" statistics. This seems to be intended to add a random element to each model created. So you could create 2 light tank models and have different stats based on how those random numbers turn out.

The stats seem to make more sense when most of the rolls are "offensive" instead of "defensive". Here is an example for an extreme case where all rolls go for the "defensive":

[all5n]

And here is the resultant comparison to the standard light tank II model:

[all5n]

And another example of all "offensive" rolls on the upgrades:

[all5n]

And the stats:

So there it could be something in the weighting of the trade-offs between offensive/defensive numbers.

[all5n]

One More. This one seems a little more in line with what i would expect.

[all5n]

All stats look higher than the base model, save the offensive vs. tanks. Which is odd since thats what the Tank Combat I and II are supposed to impact.

[all5n]

Another one with similar upgrade text, but the stats are worse on almost every mark.

[all5n]

Stats

[Vic]

Hi all,

Took the time to recheck the algorithms.

And actually everything is performing as it should be.

However I should clarify a bit.. The random model generation is random. Its possible to get worse stats than the Std. Tanks. If you create a new level I tech tank it has about 16% chance to be worse on a specific stat group, about 48% chance to be the same and 36% chance to be better. This randomization effect is increased if you create higher tech models from scratch. (level II,III,IV)

The trick here is to use a bit of strategy (and luck) with creating models. I would advice to first research 2 or 3 different models at level I tech. Then when you have 1-2 models that are better or close to the std. version you buy tank research level II and upgrade your best models. If you add in the effect of the 9 levels of extra anti-Infantry and anti-Tank research fields for the tank models the final result defenitely has a very good chance to be better than the std. versions. At a cost of political points of course.

Also keep in mind that the ATG New Dawn ruleset is basically a test bed for new features I eventually want to perfect for Advanced Tactics Platinum. The final idea for these models is to no longer have the option to chose Std. version production as well as removing the statistics from view. I just left it in for now for players who don't want to design models and for the AI.

For the technicals under you. Its possible to improve the average new model by changing event 96 line 7. A variable there is now set to 41. If you set it to a lower value you will get much better results with new models. If you set it to 0 the models will always be better.

Hope this helps a bit.

Best wishes,
Vic

[Calliope]

Thanks Vic[:D]. Excellent work.[&o]

And now the only question left is the cost performance. As we all know that all tank researches are only used by the tank units which is newly designed. Which means only a small part of troops get this buff. If you want to enjoy this new tech fully, you have to use additional PPs on leveling up raw mines and oil to ensure that you can mass them. So just like I have mentioned above, in most cases put production points on other troops or other researches are much better in this version.

Fox example, I get a new game and the rule is stone age. Here is the list of how many PPs does a level 2 light tank need(different in each map but the ratios are the same):
Research Light Tank: 43PPs
Tank Anti-Inf I&II: 10+32PPs
Tank Combat I&II: 21+65PPs
Tank Mobility I&II: 10+32PPs
New design: 21PPs

Total: 234PP (If you get a bad design there are more PPs needed)

And for comparison, in this case the Staff II research only needs 261PPs. So here the question comes, why should I use PPs on such an area that only fews of units enjoy the buff, and the buff is too small even in some cases needed to remedy the bad design. The Staff tech can benefit all my troops and the buff is obviously much more than tanks, there are also no risks(no bad design) or more investments(for tank research you still need to level up the mines and balance the logistics well etc). So my advice is that the cost of tank research should be reconsidered. If I want to enjoy a game, I can get some new tank designs for fun. But If I want to win a game, I have to consider the cost performance.[:)]

[Tac2i]

@Calliope: good analysis but if you are playing in a multi-player game and your opponent develops a mean new tank model and you avoid the feature due to cost, your opponent could win the tank war and scuttle your chances of victory. I'm not saying cost of tank modeling may not be an issue but I do know that Vic increased the cost slightly during the beta period. Obviously we need some v2.20c multi-player games under our belts to see how this works in practice. I've involved in one such game now. I'm of the opinion currently that research/new tank designs should not be cheap. If things are too cheap then everything gets upgraded by everyone and then what's the point? May as well not have research or tank modeling. I like it when players have to make decisions about what to prioritize.