Return of investment rate?

[stupidnoob]

The problem im often musing about is whether to buy a new item/tank now or later. The advantage of buying now is, that one gets more firepower/hp and therefore can kill more creeps and enemy tanks per time unit(gaining more money) and die less often and force the enemy tanks more often to heal(causing them to gain less money).
The disadvantage is that one has a slot for a cheap weapon/armor or invest in a tank, that is only useful for a short time. Thereby one loses some money.

Therefore it depends on the return of investment rate.

Now my question, what return of investment does a basic magic actually provide at least in regard to creeps. Demolisher with one upgraded basic magic and 1361 gold on middle lane. Buy basic magic or wait for swarm rockets?
How much money per second is to be expected from the extra basic magic, assuming that no additional tank kill is generated?

Since in hb the return of investment rate is doubled, in that mode i always buy if it does not require a teleport and often if it does, since it pays very quickly, but im uncertain about normal mode.

[Prog]

The problem is a bit more complex in my opinion, as the roi/sec of an item is dependent on the opponents, which makes it really hard to give a general assumption. There are situation in which you would fight at your tower without getting one additional basic magic, so the (effective) roi/second is quite huge, but there are other ones in which it has a rather low impact on your creeping. (And of course there are the possibilties of getting tank kills with additional weapons which are closely related to the opponents skill and strategy) It is of course possible to do datamining for values actually found in games and evaluate them statistically, but that's too much work for me and would have quite a low impact on a single game, as you would need to modify the results anyway according to your opponents.

[qweqqweq]

right now the -stats is screwed up for player red since he's getting credit for a lot of creep kills which aren't his. but if you play as any other player i think you can type -stats and then -stats x to check your progress with other players. if i'm not playing as red I use -stats to know if my enemies might be saving up for a tank that could crush me in an instant especially with teleport. Even then i've found it useful to just get a teleport breaker and keep saving for a teleport breaker which is a lot cheaper than a hull in some instances requiring just 3 minutes of farming more to get what i want.

there are a lot of things to consider. if you're extra early investment on dps not just gets you extra creep kills and tank kills but also wether or not it'll force enemies to sacrifice their own creep and tank kills to defend against you. for instance, if you force them to chase you around the map or to cp tp to your lane then they're loosing gold and it's allowing your teamates to farm more easily by relieving pressure off of them. it's 150 gold round trip for a cp tp's so if i can't take a cp then i try to get 3 or more enemies to come to my side lane without dying. If it's early on in the game and my bounty is still less than 300 gold even if 4 or 5 come and i do die i still consider it to be a small advantage for my team or at least not an advantage gain for them.

[stupidnoob]

The teleporting is a good point i never thought of. In early game dying gives enemy 200 gold, so if 2 have to port in and back for 300 and oneself has to port after dying for 75, it could be a good deal, depending on what allies can do in that time.

I understand, that the true rate is too dependant on a lot of factors, i was interested in the effect on money gained from creeps while just standing around, where one can hit the creeps. That sometimes happens in the mid with destroyer and the enemy having long range, my guns do little there except killing creeps.

[Dr.McNinja]

Like Prog said its complex... I think its fine to want to understand the game better with statistics but what youre asking is probably too hard and time consuming to evaluate.

You would have to test the same configuration in a similar way doing similar things to see how much 'income' a weapon or configuration gives. Therefore its all up to the player's experience. Its too difficult with all the variables such as enemies, luck, etc..

For me I have a method I dont always follow that goes (5k): single strong starting weapon > 2500 hull > another similar weapon > gold hull > strong weapons etc..
I almost never buy 1250 hull in 5k hb games. Upgraded weapon and hull can do quite well in farming up 6k+ for your next weapon, leaving you still will 3 slots or 2 if you have speed boost. If im doing bad or average, i will probably not upgrade and buy the same 3-4k weapon after getting my first hull. Sometimes when i screw up by saving too long, others outpower me by the time i get my weapon for 2 reasons, they farmed more than you; you farmed less or you died and missed kills because of basic equpiment.

Instead of talking about a 'rate' i think it can only be talk about in strategy-terms. Notice how we've only put forward potential situations rather than a rate gold/second or gold/minute.