Oh yeah, you were doing this kodachi only spam as a start before, now you must be laughing.

I still believe that this early kodachi spam being this viable is bullshit. It allows it's user to potentially win a game without much effort invested into the match - just good RPS and a bit of luck with timing (where the comm last jumped, whether a con or llt was made prior to the koda being spotted). It's basically a difficult to shut down form of lazy cheese.

UG got both me and Randy with it yesterday and will likely continue to use it. Here's to hoping a cure is found in the coming days.

I think it needs to be respected to counter it. PRO_rANDY does a greedy 2 constructor start in a fairly indefensible part of the map. This game looks like a rush beating a greedy economy start, in part because the economy start does not respond swiftly once the rush is scouted.

[Spoiler]
0:58 - Enemy raider seen on radar. The only defense at base is an incomplete Dagger. The currently scouting Dagger should probably intercept instead of continuing on.

1:08 - The scouting Dagger spots another Kodachi. This implies that at least two Kodachis are coming towards PRO_rANDY.

[Spoiler]
1:15 - Scouting Dagger lost. Commander is making a Mex on high priority. Constructors are making Solar and Mex. At this point it is surely time to make a Lotus or assist the 35% complete Dagger in the factory.

[Spoiler]
1:49 - Three Daggers lose a static battle against two Kodachi. Another Kodachi raids. Two constructors are making a Solar and a Mex while the Commander finishes up an irrelevant Lotus.

The comander has to retreat since there is now nothing that can deal damage at the base. The game is over by this point. I think it is reasonable that 340 metal in raiders beats 225 metal in raiders, especially since the cheaper army is 23% faster. A Lotus in the base at any earlier point in the game would tip the balance and make Kodachi unable to attack. Breaking the ensuing containment should be doable with Dagger.

Kodachi spam may well be too good, but this is not a game that demonstrate it. I expect a hard rush to elicit some reaction and change of plans from greedy starts. The two Kodachis at the back could just as easily have been three Socrchers.

Also, Omoi's been tweaking the kodachi prefire widget and honestly it makes dealing with them even more of a nightmare. If I hadn't done a scalpel start deliberately he may have toasted me, because daggers were unable to effectively chase anymore.

I can't really put my finger on it, but if you compare Kodachi with Pyro it just seems a bit OP.



Pyro costs 30% more.
Pyro is 23% slower.
Pyro has 13% more range which the Kodachi can somewhat make up for with greater speed.
Kodachi has 10 HP/sec regeneration after 10 seconds. Pyro has the standard 5 HP/second after 60 seconds without taking damage.
DPS is hard to compare as Pyro multihits targets. For example a Pyro will destroy a factory 2-3 times as fast as a Kodachi but against a Jack, it will only have about 10-15 percent more DPS.
Against lone Lotus turrets, a Kodachi will have like 100-200 HP left while a Pyro will have about 150-300 HP left.
So a Kodachi can go in, destroy a Lotus turret, regenerate to full health in about a minute. A Pyro will take 2-3 minutes to do the same.

Jump at max jump range gives the Pyro a little extra speed but it is still only barely faster than a Kodachi.

So a Pyro can potentially do lots of damage if it actually gets behind enemy lines but it is much less likely to get there, as it is much less likely to make it through light defenses, and enemy raiders, and any poking damage it recieve takes at least a minute to remove while a Kodachi can be poking around and regaining that health while speedily moving around.

As per my taste, it could both become more generic (lose many of the non-tank-aesthetic magical bling it has like regen or fire or fire immunity or dot etc) and be unique.

Currently no raiders have unguided arcing weapons. Of those that have arcing weapons at all, there's Duck (perfect-accuracy missile) and Cutter (same, but also it's a ship). Even if kodachi lost ground fire, sticky fire, aoe, regen, fire immunity and regen it could still stay reasonably unique simply by having an accurate ballistic shot (autocannon?) on a raider chassis.

(Glaive is technically ballistic, but extremely low gravity and lifetime of the projectiles means glaives can't shoot over each other)

Removing Koda fire also allows to remove its fire immunity, thus making Pyro vs Koda much less hilarious. Remember when making one vs the other was suicide because Pyro did damage and Koda did fire? The current koda reduces this; well, it could reduce it further, to zero. IIRC no other units except these two are fire-immune.

Godde you raise some good points. Pyro is a bit undertuned (due to its difficult balance) and I was trying to redesign Kodachi into a raider that would be used and tested. Also jumpjets are quite a powerful combat and cliff-passing ability.

Anarchid a non-AoE precise arcing shot sounds like a massive pain. Raider battles would be decided by jitters and jinks because the difference between a projectile hitting or not would be so important. As I see it, Kodachi can afford to have a slow projectile because the AoE provides a graduated success scale for projectile hits.

The slow projectile is one of the more annoying parts of this Kodachi but it is also unique.

The auto-regen is very powerful, even in a rush like this, because rushing Kodachi can be saved and return as more standard raiding and harassing force. Similar rushes with other units are more all-in because failing results in too many lost and damaged raiders.

The danger of balancing an auto-regen Kodachi is that it could only be good in rush and earlygame situations, since in higher densities more Kodachi would die. I like auto-regen for a light tank unit since tanks have to invest a lot in repairing. Perhaps the autoregen could simply be halved.

Kodachi could be slower, but it is the fastest Tank, while both Rover and Hover have faster units. It is by far the heaviest fastest unit so there is room for it to be a bit slower.

What I really want to do is leave the patch for a while and see what happens.