I’ve been playing around with a progressive snapshot, as well as googling folks who have used it. Some recurrent complaints are from folks who rack up insane numbers of hard brake events. I find it a bit hard to believe that such cases are legit.
Initially, I drove like a grandma so as not to set it off… but then looking at the raw data, I decided to keep pushing things to see where I might actually run into trouble. Even with slightly aggressiv driving, I had zero hard brake events (granted, if I still had my Z31 from years ago, that would be a different story). Apart from that, I did log a couple hard brake events during a 30 day period… namely I had stop fast to avoid a collision… and being the decel occurred in less than 3 seconds, I ended up with 2 hard braking events. Apparently progressive doesn’t count isolated events such as these, as I still got a 30% discount.
In an emergency stop, like coming around a blind corner and finding a rolled semi blocking the entire road, one is very likely to exceed a 7mph/sec deceleration. In fact, most traffic safety speed recommendations are based upon a decent driver in a well maintained car being able to decel at ~11mph/sec to avoid a such a collision. A traffic study where folks had to decel at 12mph/sec to avoid a stopped up traffic jam in a blind corner equated to a whole ton of accidents… as not everyone is decent driver, nor are all cars well maintained and, a lot more folks talk/text today in comparison than when the 11-12mph/sec standard came into being.
As far as even higher deceleration events go, it takes practice and meticulous maintenance and design. A new squad car and a highly trained driver can decelerate safely at near 21mph/sec. In F1, its even possible to get upwards of an average 45mph/sec deceleration with peak decel over double that. Such requires an initial velocity of over 170mph, carefully tuned aerodynamics, and a wealth of pristine tuning and driver training to pull off.
Bottom line, there is no question in my mind that Progressives criteria for discounts, as well as most trucking companies internal standards for critical event recorders equate hard hard braking as equal to or greater than 7mph/sec.
However, there is a ton of anecdotal evidence of folks racking up hard braking events via the snapshot… this is often followed up by said folks saying they have never been involved in a accident. While its entirely possible they really are bad drivers and have been very very lucky, the number of complaints and lack of accidents doesn’t smell right. Granted, folks often estimate their driving skills and their cars as above average… I’m not sure that’s the whole story. Certainly the older individual cited earlier who ran up 1 hard brake event for every 10 miles would have to be extremely lucky in order not to have been in a few serious accidents over the years.
Something seems amiss.
ODBII’s (the port the snap shot plugs into) original purpose was emission diagnostics… over time the diagnostic path has grown multifold. In some cars, the ODBII connector has become a defacto access port to a cars internal communication network…. in some ways, the ODBII port is not unlike an ethernet jack in your home or office. Imagine if you put a piece of monitoring gear on it, that periodically checks to see whats going on. Under normal conditions, its not a big deal. However, lets say you have a lot going on, like 3 videos playing, a couple VOIP calls, and the network is at capacity. Now, imagine what happens when you add more to the network than what it can handle… something has to give. Video slows, VOIP drop outs increase, and monitoring gets a bit intermittent.
In your car, brake computer/engine computer messages are mission critical, they cannot be usurped by diagnostics. In other words, the diagnostics get answered when time is available, which is not necessarily the time that an external device wants. The end result, is that random response delays can enter in during times of high activity. It may be possible that such delays end up causing less than reliable determinations of braking decelerations. This would of course vary from OBDII implementation by year, make, model, and possibly option package.
If such is the case for a specific year, make, and model of a car, it would seem that Progressive would be well aware of it. Not giving legit discounts to safe drivers is likely to result in lost customers which would be counter to their profit motive. I am pretty sure they would look at aggregate snapshot data to determine the potential for errant data is by year, make and model, so I think this possibility is pretty remote.
There are however 2 areas where Progressive is unlikely to have enough information to make the call as to network capacity and thus hard braking error rates. Vehicle option packages and current vehicle condition, both of which may impact ODBII network traffic negatively. Ie, if there is noise on the network as something is wearing out… network retries will go up substantially and bandwidth will be lost. In a similar vein, comfort control options packages and/or concierge services may also use the network and thus reduce available bandwidth.
Bottom line though, all of this is just a wild guess. Short of hooking a ODBII network analyzer/sniffer (which is not an off the shelf type of test gear you could get at auto parts store or have run at the neighborhood garage) there is no real way to know. It might well be there are significant numbers of bad drivers who are also incredibly lucky.