Wednesday, July 23, 2008

NCAA 09 (360): Gameplay Discussion #2

Today, let's delve deeper into the specific gameplay of NCAA Football 09.

[please note: when I refer to slider settings, those settings are the same for both Human and CPU, because right now, the sliders are broken. If you change the Human sliders, they change the CPU sliders as well.]

Based on my own observations after playing the game for a little over five hours, I identified the following areas as "broken":

Special teams:
--kickoff and punt coverage

--CPU quarterback completion percentage (way too high)
--CPU quarterback pass selection (almost no deep passes)
--CPU quarterback response to blitz ("psychic" power)
--CPU quarterback (ineffective as runner)
--CPU running game (ineffective)
--CPU running backs (overuse of jukes and special moves)

--CPU defensive line (unable to pressure quarterback)
--CPU defense (unable to play man-to-man coverage)
--CPU defense (pursuit angles totally broken)

--penalties (almost none called)

If you're wondering if that represents all the gameplay--yeah, it pretty much does. Essentially, it's all broken. And I'm not even discussing rare situations that only happen occasionally--these are every-game occurrences.

To demonstrate what's happening, let's look at a classic matchup: Florida versus LSU in Death Valley. Heisman difficulty. Again, this is in CPU vs. CPU mode, which I use to document issues I see when I play the game.

Florida won 24-7, in case you're interested. Now, let's look at the numbers.

There were 91 total offensive plays in the game (that doesn't include special teams plays), which is low, but for our purposes, it's more than enough. Let's review the gameplay list in order.

Special teams:
1. Here's how far away the nearest tackler was when a punt returner caught the ball: 33 yards. Oh, except when it was 40 yards, which happened sometimes as well. 33 freaking yards. Really, that's so bad that it's stunning.

1. CPU completion percentage: there were 52 passes in the game, and 36 were completed, for a 69% completion rate. That's with the QB accuracy slider reduced to 45 from the default of 50, and the "catch" slider at 0.
2. Deep passes: Of those 52 passes, 3 were thrown more than 20 yards beyond the line of scrimmage, or 5.7%. This is with the fastest wide receiver in the game playing for Florida.
3. CPU QB "psychic response" to pressure: I don't collected data for this, but it's amazing to see how often a quarterback under blitz pressure will find the open receiver and throw a strike. Almost every time, seemingly.
4. CPU QB running: Tim Tebow, the best running quarterback in the country, had 9 rushing attempts for 11 yards. His speed is rated at 85, which makes him absolutely useless, because he's running so slowly that he can't get into the open field.
5. CPU running game: 39 total rushing attempts for 85 yards, or just over 2 yards a carry. There will be games where you see much higher rushing totals, but they'll always be from one or two 60+ yard runs. There's no such thing as a consistent ground attack, at least not what from what I've seen.

Oh, and that's with run blocking at 80 and running back ability at 75, by the way. Sliders aren't going to fix this.
6. CPU jukes and special moves: another observational point. It's remarkable to see how many special moves the CPU running backs will make when a defender isn't even close enough for it to matter. This includes jukes that actually take them out of bounds (oops).

1. Defensive line pressure on passes: with 52 passes in the game, let's assume that half of those pass plays were facing a 4-man defensive line, while the other half faced a 3-man line (that's conservative, because I think most of the passes were against a 4-man line). In other words, there were 3.5 blocks on defensive lineman per pass for 52 attempts, or a total of 182 on-one-one situations between an offensive and defensive lineman in pass situations. How many of those confrontations resulted in a defensive lineman defeating the offensive lineman and running free toward the quarterback?

5 out of 182 times, or 2.7%. With the four starting defensive ends for the two teams rated at 95,92,89, and 89 overall.

Think sliders can fix this? That's with Pass Blocking at 0 and Break Block (for the defense) at 100.

Now a knockdown would count as pressure, because it results in an incomplete pass. I put the Knockdown slider at 100, and there were 2 knockdowns out of 52 passes.
2. This is observational, but on almost every occasion where I saw a defense playing a man-to-man defense, they got burned. With the speed differences so exaggerated this year, it's money.
3. Defensive pursuit angles are comical this year--there's no other proper word to describe it. Here are two videos that show it perfectly. In this link, go to post #322 on the page and watch:
I don't want to tackle you

That was bad, but this is EPIC. Go to post #371 and watch this:
we don't want to tackle you

I'd describe those videos, but they are so incredibly embarrassing that you need to see them for yourself. Really, they're staggering. There are many times in NCAA 09 were the pursuit angles look like they were copied from Ten-Yard Fight. Seriously.

1. There were 2 penalties called in the entire game. That's with all penalty sliders at 100.

See what I mean? We're not talking about minor problems with gameplay here--we're talking about major, fundamental problems across the board.

That's what happens when you ship an early beta, because no one could play this game for more than a couple of hours and think it's anything more than that.

If you think I cherry-picked two teams and set this up just to get the numbers I wanted, you would be incorrect. Maybe you'll get lucky and occasionally see a game where the CPU runs effectively, or you'll see 6 penalties called (which is still too low), but you'll see all of these problems in the vast majority of games that are played. These issues aren't rare flukes--they're endemic.

Tomorrow: don't even think I'm done yet. It takes multiple days to document everything that's broken in this game.

Site Meter