Earlier this month, I posted a suggestion for hiding player ratings until the player has played enough games to reveal them. Afterwards, I posted a list of my suggestions for offense, defense, and special teams for Madden 18. That still leaves some other outstanding areas of improvement such as Franchise mode, and I'd like to spend this post focusing there.

Let's start out by going over some of the things that are left over from last year's wishlist:

• film study
• weather forecast
• preseason substitution logic
• hot blocks
• delayed kneel down and rugby punt
• onside kicks and loose ball recovery
• coach's challenge option
• scheme experience
• training camp
• multi-week scouting
• Mulligan the previous draft
• time to read the opponent's offense

A lot of these items are related to Franchise, and so keep them in mind as you read through this post. But before I jump into franchise suggestions, let's first look at the issue of the Accelerated clock as it has been implemented in Madden for years:

Accelerated clock, two-minute drill, and CPU timeouts

I've brought this up before, but clock management really needs to be addressed. The accelerated clock should never be disabled! Not in the two-minute drill; not ever. The two minute drill is when it is most important to enforce the accelerated clock because otherwise it completely breaks the two-minute drill. The CPU is particularly bad at exploiting this. I regularly see the CPU go into a huddle and break it within 5 seconds of game clock, which is faster than if they had tried a hurry-up, and which spares them from using a timeout. Human users can exploit this same tactic as well in order to avoid the time it takes to run up to the line. All you have to do is quickly select any pass play then audible or hot route your receivers.

Breaking a 2-minute drill huddle with 35 seconds on the play clock while
the game clock is running completely breaks the 2-minute drill.

And speaking of CPU timeouts: the CPU should actually use them. There should be some logic in place where if a CPU QB either can't figure out the pre-snap coverage, or he doesn't like the pre-snap coverage, then he should call a timeout to mulligan the play. This should happen if the defense puts eight men in the box when an inside run was called, or if the CPU QB reads press coverage on a wide receiver screen, or other such situations in which the the CPU determines that the player's play is likely to trump their play call. This should be tied to a QB's Awareness and/or Play Recognition rating, and there could maybe also be a small chance that if the QB doesn't make the right read, then the coach can call a timeout instead. CPU defenses should similarly be able to burn a timeout if they read a particularly unfavorable personnel match-up.

Another improvement that could be made to the accelerated clock is to add some variability to it...

I recently wrote regarding a proposal for improving practice squad and training features by hiding player ratings until the player has played enough games to reveal them. Of course, there's still a lot of other aspects of the game that I'd like to see improved. I was pleased that this year's Madden 17 implemented some items from my wishlist from last year. There's still a lot from last year's wishlist that I'd like to see implemented in some fashion. Playing Madden 17 has also raised new ideas for improvement.

Let's start out by going over some of the things that are left over from last year's wishlist:

• film study
• weather forecast
• preseason substitution logic
• hot blocks
• delayed kneel down and rugby punt
• onside kicks and loose ball recovery
• coach's challenge option
• scheme experience
• training camp
• multi-week scouting
• Mulligan the previous draft
• time to read the opponent's offense

Now, admittedly, a lot of the following suggestions are going to be based on my own subjective experiences with the game. And these opinions come from someone who is almost exclusively a single-user Franchise player. My priorities are going to be far different from the desires of MUT players or even online franchise players.

I'm also not going to bother (right now) with the obvious problems: rubberband AI that creates obnoxiously artificial "momentum swings", the broken man coverage, robo QBs, the complete unwillingness of my linemen to block at the point of attack on run plays, or the down-tuning of new features (such as throw out of sack, aggressive catch, and defensive line moves) to the point of irrelevance, and so on. Instead, I'm going to try to focus on less-obvious mechanics that interact with these problems and which have forced EA to make the [bad] decisions that they've made.

Better run-pass balance and longer games

The general design of Madden isn't very run-friendly. The fact that the game is balanced and tuned for quick, 6-minute quarter, pick-up-and-play online matches (instead of full 15-minute quarter games) means that grinding it out on the ground to establish the run is futile. Trying to run the ball in a 6-minute quarter game (with accelerated clock turned ON, which is the default) can rapidly burn through time. I regularly eat up an entire quarter and a half in a single drive when I commit to the running game in such matches, and that is just unrealistic. This forces both players and the CPU to depend on the passing game to score before a half expires. In my opinion, this is a fundamental design flaw of Madden, and the game will never be truly great as long as 6-minute quarters is the focus of design.

Madden's fundamental design is not very run-friendly.

But fundamental design flaws aside, my experience with Madden 16 and 17 has been that the CPU is completely inept at running the ball. Even when the blocking is solid, the CPU-controlled back can rarely identify and hit the hole, and usually runs right into a waiting defender or one of his own blockers. CPU backs are even worse at running to the outside, as they'll often run backwards in a futile attempt to get to the edge, instead of just cutting upfield for whatever yardage they can get. This often leads to large losses of yardage, backs up the CPU, and contributes towards the CPU's over-reliance on passing the ball. I usually play with the CPU Run Blocking A.I. slider up between 80 or 100, and yet CPU running backs still routinely finish games with stats along the lines of 15 rushes for 20 total yards. Pathetic. If a CPU runner does have a successful game, it's usually because they broke one or two long runs due to a missed tackle, and 90% of their yardage total comes from one or two plays. Also pathetic. Seriously, I have rage-quit an hour into games because of the CPU's ineptitude.

CPU Doug Martin runs right into his pulling guard [LEFT] instead of going inside like the trap play is designed.
CPU Doug Martin has a huge hole with only a single cornerback to beat [RIGHT], but cuts into traffic instead.

I also have a lot of trouble running the ball myself with my own Run Blocking A.I. slider set to anything below 60 or so...

Well, here we are at the end of the NFL season already. Seventeen weeks are in the books, and the 2016/2017 playoffs are in full swing. As we prepare to say goodbye to the 2016 NFL season, it's also time to start looking ahead to the future of the Madden video game series.

I was really happy to see the practice squad and weekly training added to this year's game. I feel that this allows for more realistic development of players over the long term, and it makes the draft feel more worthwhile since you no longer have to cut your late-round draft picks. Despite being a good thing to have, the practice squad feature has some problems.

Being able to see all of a player's ratings make it trivially easy to poach other team's practice squad players.

For one thing, determining who to start and who to throw on your practice squad is a pretty trivial process of comparing numbers in a spreadsheet. Heck, you can usually get away with just comparing a single number: their overall ratings. There are no "judgement calls" to be made here.

Practice squad poaching is another problem. Any player with a rating above 70 is likely to get poached off of your practice squad by another user (even by CPU teams). The reason for both these problems is that it's trivially easy to know how good any given player is - the game shows all their ratings right there in the menu. You don't need to put either of them head-to-head in practice or put them on the field to see how they perform. The ratings dictate performance, and the ratings are publicly available.

Uncover rookie ratings during training camp

How can we resolve this problem of practice squad poaching? Well, we can hide rookie ratings until you actually practice with them and play them in games. Much like how the true ratings of players in the college draft are hidden until you scout and/or draft them, the game could also hide the true ratings of incoming rookies.

This opens the possibility of a training camp feature being a valuable tool for player assessment. I've already proposed a training camp feature in my previous wishlist, but this idea could supplement that. As you put players through your training camp, you'd slowly uncover their true ratings by performing various Skill Trainer drills or other practice activities and scrimmages. Then, once the season begins, you would reveal further ratings through weekly training and by playing the players in actual games. This would also have the effect of adding further value to preseason games, as you'd use those as a proving ground to hopefully uncover any remaining key ratings for your young players. You'd actually have a genuine reason to play them in games because you honestly wouldn't know how well they'd perform.

Perhaps the ratings of incoming rookies should remain hidden, even after they are drafted?

Any ratings that you unlock would remain hidden to all other teams, so that they won't be able to simply compare overall ratings with their own players. Each team could then maybe have the ability to spend some of their college scouting points on scouting other teams' practice squads looking for players that they could poach. Doing so would gradually unlock some practice squad player ratings.

There could also be a set of publicly-known ratings for each player that would be known to all teams in the league. These would be unlocked as the given player plays in games, and playing in nationally-televised games (such as Monday night) could maybe even accelerate the unlocking of ratings. So players who have been in the league for a long time, and who have lots of public game film would be more of a known quantity. We would all know how good Tom Brady is, but we wouldn't necessarily know how good Jacoby Brissett might be until he actually plays some games.

The entire NFL knows that Tom Brady is a superstar, but not as many people know if Jacoby Brissett is any good.

In the meantime, the game could provide some more "fuzzy" ratings for players whose true ratings are unknown. Either keeping the grades that are used from college scouting (A, B, C, D, etc.), or by providing ranges for unknown ratings (e.g. a player's catch rating is between 75 and 85)...

I've been on a few podcasts over the years. I had a handful of sessions on the now-defunct Geek Fights, and I've been an almost annually-recurring guest on the Civilization podcast PolyCast. Now I've got another feather in my internet nerd hat, as I've recently done a guest appearance for On the Branch Gaming, a fairly new video game Let's Play and commentary channel on Youtube.

The people who create this show are actually personal friends of mine, and I've been meaning to participate on the show with them in the past, but scheduling conflicts made it difficult for me to do so. But since I had lots of free time over my extended winter break, they invited me on to play one of my personal favorite games, Dark Souls!

At the risk of seeming like a bit of a jerk, I had never actually gotten around to watching any of their previous episodes. I had subscribed to their channel (which they insist is the important part), but I was coming into this recording completely blind, not knowing what to expect. As such, my appearance may provide a somewhat different feel to the show. Hopefully it is well-received. We all certainly had fun recording and watching John get his butt handed to him by the legendary Dark Souls difficulty.

In fact, he had to spend some time out of recording grinding and leveling so that he could actually get to the game's first proper boss (the Taurus Demon) in a timely manner.

You can watch the entire first episode below:

I hope you enjoy the video. If you do enjoy, please subscribe to On the Branch's channel and check out their other videos!

Perhaps you'll see me again in future videos. Perhaps John will "git gud" and brave the dangers of Lordran with me at his side again? Or perhaps maybe I'll convince them to play Silent Hill...?

Cities: Skylines has always been a game that takes some different approaches to city builder gameplay. The base game explored how a city's geography can influence the development of the city, and also put a particular emphasis on designing efficient transit infrastructure by allowing the player incredible freedom to construct your own roads, highways, and interchanges, rather than relying on prefab ramps and over/under passes. None of these concepts were new to city builders, but Skylines added nuance to them and made them much more active elements of gameplay.

Its newest expansion DLC, Natural Disasters, follows suit. This is a very difficult expansion to review because its content - by its very nature - is random and unpredictable. Natural disasters aren't new to city builders. Disasters were a popular component of the old SimCity games, as many players enjoyed building up their beautiful metropolises only to unleash earthquakes, tornadoes, meteor impacts, and even alien invasions and dinosaur attacks and watch it all burn. Now Skylines has support for this fan-favorite SimCity feature, but it takes this commonplace feature in some new and interesting directions.

Installing early warning and detection systems will give you advance notice when a disaster is imminent.

Most notably, Cities: Skylines' take on disasters puts emphasis on preparation for disasters, rather than on the chaos of the disaster itself and the clean-up in its aftermath. Like with its SimCity forebears, disasters are something that you can toggle on or off in the game's menu, and you can also adjust their frequency. When enabled, you'll encounter disasters of various flavors ranging from forest fires to lightning strikes, to tsunami and meteor impacts. You'll have to make sure that your city is protected by preventative measures, and that it's protected against these eventualities.

Early-warning systems like firewatch towers, weather radar, and space telescopes can warn you about forest fires, storms or tornadoes, or even incoming meteors (respectively), and can mean the difference between your citizens having enough time to evacuate, or half your population being buried under rubble. You'll need evacuation shelters for you citizens to escape to, and each shelter needs to built long enough in advance for it to be stocked with supplies of food, water, and other essentials (which must be pulled from your city's economy or imported). And lastly, you'll need radio towers to warn your citizens to get to their designated shelters.

Building emergency shelters, stocking them with supplies, and planning evacuation routes will protect your citizens.

You can also create planned evacuation routes similar to bus routes that will pick up residents and take them to a shelter. I had some trouble getting these routes to work properly though. The fact that the buses are dispatched when an evacuation is activated means that they often create a log jam on your roads as they all funnel out of the shelter. I also had issues with the buses apparently not picking up people who were at more distant stops on the route, since the areas along the route that were far from the shelter never managed to evacuate. Maybe there's some trick to getting these routes to work properly that I just haven't found yet. But this does highlight one problem with the expansion: its new systems are not very well documented or explained...