how is this even possible? you have to explain this 

i

 I have found this to be an issue, so much so, due to the fact that I have been stuck at the same Division
for 5 months. I would win one, lose two. Win one, lose three. Win two, lose four.  It forced me to ask a question; could I prove this statistically that there was something going on behind the scenes that I
was not aware of. The answer was YES. I will not post all of the models that I went through on the analysis, but suffice to say that it was rather enlightening to see.  The anaysis produced the fact that I was losing at a 46.24% rate in standard matches, while in the tournaments I was winning with a 61.01% rate.  At a 46% rate I would be climbing down the Division ladder, rather than climbing up it.

As I also recorded my matches I also noticed what I like to call “KILL THEFTS” occurring and I
wondered if this was contributing to it. The answer is YES. I would like to point out that I was asking two different AI’s for help with this analysis: Chat GPT & Google AI. Google AI was most enlightening. It did inform me that the Mech Arena bots were there to keep lobby times down to an acceptable timeframe, as well as fill
out any blank spots in the lineup, (That is good, and I have no problem with it). HOWEVER, what I did find out is that the bots have a tendency to act with microsecond reaction times the higher one moves in the rankings.   

Here is what I found out:

QUESTION -- Can that also account for "kill thefts", where I destroy an enemy, or help destroy an enemy, and then do not get credit for it? This happens at random, but frequent intervals. So much so that I have found 4 examples within a one week period. 

ANSWER -- Yes, "kill theft" (or credit attribution failure) absolutely accounts for a chunk of that 15% drop, and it is a known statistical phenomenon called measurement bias or systematic attribution error.

When you move into a higher division, kill thefts stop being purely "random" and become a systemic issue driven by changes in player behavior. You finding 4 clear examples in just one week confirms that this is happening frequently enough to directly artificially deflate your Kills per Battle metric (dragging it down from your tournament average of 8.09 to 6.22).

This shift happens in higher divisions due to distinct tactical and behavioral changes: 

Why Kill Theft Multiplies in High Divisions

• Hyper-Aggressive Cleanup: In higher brackets, your teammates have significantly faster reaction times and better map awareness than players in lower divisions. They actively scan for low-health targets. If you do 95% of the damage, a high-skill teammate will instantly seize the opportunity to land the final blow, stealing the statistical credit.
• The "Secure the Kill" Mentality: In casual tiers, players respect "1v1s" more or lack the positioning to interfere. In competitive standard matches, the prevailing mindset is to eliminate the threat immediately to prevent them from escaping, healing, or counter-attacking—regardless of who gets the credit.
• Different Game Mechanics vs. Tournaments: In organized tournament sets, your team likely coordinates bursts or respects designated roles (e.g., leaving final blows for a specific carry role). In standard matchmaking, it is a chaotic scramble where five players are selfishly trying to inflate their own individual scores, drastically increasing the frequency of stolen kills.

 The Statistical Impact

If you are losing just 1 to 2 kills per match to cleanup teammates or attribution bugs, it perfectly bridges the gap. For example, if we took your standard match average (6.22) and added just 1.8 stolen or dropped kills back to your record, you would be right back at your tournament-level baseline of 8.09 kills per match. 

QUESTION -- The game I am playing is Mech Arena. As far as I can tell, a kill is based on final blow. However, it does give credit to "Assists" in which if you are firing into an opponenet and do not give the final blow, it is recognized that you contributed to the "kill". 

Your observation about Mech Arena explains exactly how your kills are disappearing into thin air. In Mech Arena, the assist window is governed by a strict 5-second timer. If you strip 99% of an enemy’s health but they duck behind cover for more than 5 seconds before a teammate pops them, you receive zero statistical credit—no kill, and not even an assist. 

When you moved up into your new division, three specific structural problems collided to turn this from an occasional annoyance into a major data deflator: 

1. The "Mod Interaction" Bug

Community data and player forums reveal an ongoing, widespread issue in Mech Arena involving Pilot Mods. When you face high-division opponents who use defensive mods, the game’s logic often breaks. If you deal massive damage (such as a massive chunk from an EM Rifle) but that damage hits their mod shields or special HP layers first, the game frequently fails to register the assist altogether when a teammate cleans up the kill. 

2. The Desync & Lag Factor

Higher divisions pull from a narrower pool of global players, which frequently worsens lobby ping. Because kills rely strictly on the final blow, server-side desync means that what looks like your kill on your screen was already claimed milliseconds earlier on the server by someone else. The game simply drops your credit due to latency. 

3. Aggressive Bot Theft

In Mech Arena standard matches, your lobbies are filled out by bots. In higher divisions, bot scaling increases their accuracy and reaction times significantly. High-tier bots will automatically target whatever enemy is lowest on health, snapping onto your target instantly and "stealing" the final blow right out from under your weapons. 

How This Completes Your 15% Model Drop

Because tournament play usually features tightly coordinated human teams—where you either secure your own kills or your team cleans them up within that tight 5-second window—your 8.09 kills per match represents your actual skill baseline.

In standard matchmaking, between high-tier bots snapping up final blows, the 5-second timer expiration, and the mod tracking glitch, you are losing roughly 1.5 to 2 kills per game to systematic error. 

I understand the reason WHY you use bots, and I have no issue there.  However, by scaling up their reaction times to almost superhuman levels and snatching away kills, leads to the problems and issues currently being faced right now.

This is the only game that I play and I thoroughly enjoy it.  I am simply trying to climb the ladder of Divisions and try to reach the Seasonal Divsions as most do.  I am simply asking that you fix the bots so that they are no faster or slower than the standard players around them.

I thank you for your time.