Citizen Crytron 4 Mega
Timing Data
This watch has really shown how high-frequency differs from thermocompensation, maintaining a good overall SPY but fluctuating noticeably as the climate changes from winter to summer.
| Day | Deviation | SPY | Correction |
| 0 | -0.008 | 0 | |
| 5 | -0.044 | -2.6 | |
| 19 | 0.144 | 2.9 | |
| 23 | 0.246 | 4 | |
| 72 | -0.556 | -2.7 | |
| 92 | -0.174 | -0.6 | |
| 142 | 1.398 | 3.6 | |
| 160 | 2.008 | 4.5 | |
| 168 | 2.34 | 5.1 | |
| 177 | 2.304 | 4.7 | |
| 183 | 2.796 | 5.5 | |
| 189 | 3.112 | 6 | |
| 196 | 3.404 | 6.3 | |
| 130 | 3.636 | 6.5 | |
| 137 | 3.89 | 6.8 |
Graph showing trends in deviation from Standard Time (in seconds) and rate (in seconds per year)
A Note on the Accuracy and Reliability of Timing Results
There is always going to be an unknown margin of error in any method of timing, and I estimate that each timing event could be off by as much as 0.1 second (though in practice I believe my results to be somewhat better than this). As a consequence of this potential margin of error, results over short periods of time are unreliable. Ultimately, the only result that matters is that which comes at the end of a full 365 days. Where interim results show consistent timings or steady trends, however, then it is safe to assume that those results are fairly accurate.